fciOLOSY b G INFECTION AND RESISTANCE AN EXPOSITION OF THE BIOLOGICAL PHENOMENA UNDERLYING THE OCCURRENCE OF INFECTION AND THE RECOVERY OF THE ANIMAL BODY FROM INFECTIOUS DISEASE BY HANS ZINSSER, M.D. Professor of Bacteriology at the College of Physicians and Surgeons, Columbia University, New York. Formerly Professor of Bacteriology and Immunity at Stanford University, California WITH A^CHAPTER ON COLLOIDS AND COLLOIDAL REACTIONS BY PROFESSOR STEWART W. YOUNG Department of Chemistry, Stanford University gorfe THE MACMILLAN COMPANY 1914 COPYRIGHT 1914 BY THE MACMILLAN COMPANY Set up and electrotyped. Published October, 1914 TO a. z. THIS BOOK IS AFFECTIONATELY DEDICATED BY HIS SON 292993 PEEFACE INFECTIOUS disease, biologically considered, is the reaction which takes place between invading micro-organisms and their products, on the one hand, and the cells and fluids of the animal's body on the other. The disease is the product of two variable factors, each of them to a certain extent amenable to analysis, and it is self-evident that no true understanding of this branch of medicine is possible without a knowledge of the biological principles which laboratory study has revealed. For the purpose of helping to render such knowledge easily ac- cessible this book was written. While it is hoped that it may prove useful to the practitioner and laboratory worker, it is intended pri- marily for the undergraduate medical student. To many it will seem that the subject in general and our method of treatment espe- cially are too technical and difficult for this purpose. Our own ex- perience contradicts this. During the past three years the writer has had the opportunity to deliver lectures and to give laboratory courses on this subject to medical students of 2d, 3d, and 4th-year classes at the Stanford and Columbia Universities. It has been a pleasant experience to find the medical student eager for the opportunity to obtain this knowledge and, under the present increased require- ments for preliminary training at our best schools, fully capable of assimilating it. It is not a good plan to attempt too extensively to simplify material that, in its close analysis, presents complex phe- nomena and intricate reasoning. For this reason no attempt has been made to write an A B C of immunity as a quick road to com- prehension. No true insight into any branch of medicine or, for that matter, into any other science, can be attained without a certain amount of labor; however the concepts of this subject are, indeed, relatively simple after the first principles have been mastered, and the writer has attempted, therefore, at the risk of seeming pedantic in places, to treat the subject critically, separating strictly those data which may be accepted as fact from those in which legitimate differences of opinion prevail. As far as was feasible every chapter has been written as a sepa- rate unit. This has necessitated occasional repetition, but, it is hoped, will add considerably to clearness of presentation in each indi- vidual subject. Thepries have been discussed with as little prejudice as the possession of a personal opinion in many cases has permitted. vii viii PREFACE The chapter on Colloids was written especially for the book by Prof. Stewart W. Young, of Stanford University. Since so many analogies between serum reactions and those taking place between colloidal substances generally have been observed, it has seemed best to devote this chapter entirely to the elucidation of the principles governing colloidal reactions, so that its contents may be utilized as explanatory of the many allusions made to colloids in the rest of the text. All available sources of information have been freely used. In the large majority of cases we have had access to the original papers and monographs. However, we acknowledge much aid from care- ful reading of the admirable summaries, written by acknowledged authorities, in the works edited by Kolle and Wassermann, and by Kraus and Levaditi. Similar acknowledgment is made to equally important sources in Weichhardt's Jahresbericht, the Bulle- tins of the Pasteur Institute, and in such text-books as those of Paul Theo. Miiller, Emery, Adami, Gideon Wells, Marx, Dieudonne, and others. It is needless to acknowledge the use of such classics as that of Metchnikoff or of the many critical writings of Bordet and of Ehrlich — masters who have helped to shape the thoughts of all men working in this field. The writer takes pleasure in acknowledging many helpful sugges- tions from his associates, Drs. Hopkins and Ottenberg, and much aid, in the verification of references, from Mr. Walter Bliss, Fellow in the Department of Bacteriology. CONTENTS PAGE CHAPTER I. — INFECTION AND THE PROBLEM OF VIRULENCE .... 1 Scope of subject. Conception of infection. Attributes of pathogenic microorganisms. Forms of infection. Influences of biological adapta- tion. Classification of parasites on the basis of invasive properties. Factors which determine the power to invade. Fluctuations in viru- lence. How microorganisms defend themselves against destruction. Serum fastness, arsenic fastness, capsule formation, inagglutinability, etc. Eesistance to phagocytosis. Development of offensive properties on the part of bacteria. Specificity of different infections. Chronic septicaemia. ' ( Sub-infection. ' ' Selective lodgment in tissues. Locali- zation and generalization. Incubation time. CHAPTER II. — BACTERIAL POISONS 28 Part played by bacterial poisons in clinical manifestations. Pto- maines. Importance of ptomaines in disease. True toxins or exo- toxins. Endotoxins. Chemotactic bacterial extracts. Basic proper- ties of true toxins. Other substances biologically similar to them. Analogy to enzymes. Snake venoms. Incubation time of toxins. Conception of antitoxins. Work of Vaughan. Kesearches of Fried- berger. Absorption of toxins. Selective action of toxins. Distribu- tion of tetanus poison. Causes underlying selective action in general. Injury done during the excretion of toxins. Union of toxins with susr ceptible cells. Importance of cell lipoids. CHAPTER III. — OUR KNOWLEDGE CONCERNING NATURAL IMMUNITY, AC- QUIRED IMMUNITY AND ARTIFICIAL IMMUNITY .... 49 The struggle between the infectious agent and the defensive forces of the body. External defenses. Skin secretions. Natural vs. arti- ficially acquired immunity. Species immunity. Racial immunity. Dif- ference between individuals. Inheritance of natural immunity. Im- munity resulting from an attack of the disease. Jenner and smallpox. Pasteur's wTork with chicken cholera. Active immunization. Passive immunization. Pasteur 's studies on anthrax. Different methods of con- ferring active immunity. Methods of obtaining bacterial extracts. De- velopment of our knowledge of passive immunization. Early attempts. Behring and his collaborators. Ehrlich's work on ricin. Snake venom. Specificity. CHAPTER IV. — THE MECHANISM OF NATURAL IMMUNITY AND THE PHE- NOMENA FOLLOWING UPON ACTIVE IMMUNIZATION ... 78 Investigations on problems of inflammation. Metchnikoff 's earlier studies. Concentration of attention upon the properties of the blood. Grohman's wrork. Early opposition of cellular and humoral points of view. Buchner. Nuttall. Earlier arguments brought forward by the two schools. Behring 's summary of the situation at this time. Phenomena following upon active immunization. Earlier theories. Exhaustion theory. Retention theory. Alkalinity theory. Osmotic theory. Discovery of specific antibodies by Behring and collaborators. Ehrlich's study on ricin. Antitoxins. Pfeiffer's discovery of lysins. Agglutinins. Precipitins. Opsonins. Tropins. Conception of anti- bodies as a whole. Generalization of the facts discovered in the case ix x CONTENTS PAGB of bacteria. Haemolysins. Cytotoxins. Haemagglutinins. Precipitins to unformed proteins. Conception of an antigen. Nature of antigens. Analogy of immunity with drug tolerance. Origin of antibodies. CHAPTER V. — TOXIN AND ANTITOXIN . 104 Nature of the reaction. Earlier views. Calmette's work on snake poison. Filtration experiments. Morgenroth's work on HC1- toxin modi- fications. Ehrlich 's ricin neutralization. Development of the neutrali- zation ideas by Ehrlich and Behring. Conception of antitoxin unit. Instability of toxin. Ehrlich 's experiments. The conceptions of M.L.D., L0 and L+ doses. Discrepancy between L0 and L+. Toxoids and toxons. Method of partial absorption. Toxin spectrum. Opinions of Arrhenius & Madsen. Bordet 7s opinion. The Danyz effect. THE SIDE CHAIN THEORY. Early work of Knorr. Ehrlich 's analogy of cell with chemical substances. Analogy with ferments. Weigert's law of overcompensation. Antibodies and cell receptors. Theoretical con- clusions drawn from work with tetanus poison. Importance of lipoidal substances in brain tissue. CHAPTER VI. — BACTERICIDAL PROPERTIES OF BLOOD SERUM, CYTOLYSIS AND SENSITIZATION 134- The phenomenon of bacteriolysis and the bactericidal effect. Haemo- lysis. The mechanism of cytolysis. Amboceptor or sensitizer? The complement or alexin. Iso-antibodies. Discussion of views of Ehrlich and Bordet. Multiplicity of antibodies. Multiplicity of complement or alexin. Anti-antibodies. Neisser and Wechsberg phenomenon of complement deviation. Quantitative relation between complement and amboceptor. Congiutinins. CHAPTER VII. — DEVELOPMENT OF OUR KNOWLEDGE CONCERNING COMPLE- MENT OR ALEXIN. COMPLEMENT FIXATION ..... 168 Origin of alexin. Microcytase and Macrocytase. Anti-lysins. Alexin in O3dema fluids, etc. The question of alexin in the circulating blood. Alexin and the thyroid. Alexin and the liver. Chemical nature of complement and alexin. Cobra-lecithid. Enzyme-like nature of alexin. Filtration of alexin. Complement or alexin splitting. Return to activ- ity of inactivated alexin on standing. Inactivation by shaking. ALEXIN FIXATION. Bordet-Gengou experiments. Theoretical explana- tion of these facts. Albuminolysins of Gengou. Alexin fixation by precipitates. Views of earlier writers. Nicoll 's view. Writer 's opinion. Conception of ' ' Bordet-antibody. " Nonspecific alexin fixation. Impor- tance of lipoids. Fixation by unsensitized cells, substances in sus- pension. Anticomplementary properties ,of serum. CHAPTER VIII. — PRACTICAL APPLICATIONS OF COMPLEMENT-FIXATION METHOD. THE WASSERMANN REACTION 198 Historical. Early work on monkeys. First use on human beings. Theories of the Wassermann reaction. Methods of preparing antigen. Titration of antigen. Titration of haemolytic sensitizer. Alexin titra- tion. Performance of the test. Noguchi's modification. Modifica- tions of Bauer, Stern and others. Results of test. Reliability. Spinal fluid, etc. COMPLEMENT OR ALEXIN FIXATION FOR THE DETERMINATION OF UNKNOWN PROTEIN. Neisser-Sachs method. Principles of the method. Performance of test. COMPLEMENT-FIXATION TESTS IN DIAG- NOSIS OF MALIGNANT NEOPLASMS. Historical. Von Dungerm's method. Results obtained. Complement-fixation in glanders. Complement- fixation in gonococcus infections. CHAPTER IX.— THE PHENOMENON OF AGGLUTINATION 218 Discovery. Applications of clinical methods. Clinical usefulness. Re- lation to motility. Passive role of bacteria. Bordet 's discovery of CONTENTS xi PAGE the importance of electrolytes. Nature of agglutinogen. Alterations by heat. Alterations in agglutinability. Seasons for agglutinability. Specificity. Biological relations between bacteria parallel to agglu- tinins. Castellani's method of absorption. Normal agglutinins. Agglutinoids. Inhibition zones. Bordet's views. "Two phase" theory. Physical interpretation. The work of Neisser and Friedemann. Acid agglutination. Iso-agglutinins. CHAPTER X. — THE PHENOMENA OF PRECIPITATION 248 Discovery. Bacterial filtrates. Expansion of principle to proteins in general. Nature of precipitinogen. Specificity. Quantitative rela- tions in the reaction. Practical uses. Nuttall's studies on precipitins and historical relationship. Forensic uses of the test. Performance of the test as advised by Uhlenhuth. Influence of heat upon precipi- tinogen. Organ specificity. Ehrlich's view of the nature of the reac- tion. Physical views of the reaction. Presence of precipitinogen and precipitin in same serum. Analogy with colloids of known constitution. CHAPTER XI. — PHAGOCYTOSIS. CHEMOTAXIS. ...... 272 Early investigations. Metchnikoff 's first studies. Phagocyto- sis in lower animals. Its significance. Importance in the de- velopment from the larva to the adult. Its importance in resorption of degenerated cells. Varieties of phagocytosis. Giant cells. Leucocytosis in response to the presence of bacteria. In the peritoneum. Phagocy- tosis in tuberculosis. CHEMOTAXIS. Botanical studies. Early studies of Leber. Early studies of Buchner. Methods. Theories of chemo- taxis. Importance of surface tension. CHAPTER XII. — PHAGOCYTOSIS, Continued. THE EELATION OF PHAGOCYTOSIS TO IMMUNITY 296 Opsonins and tropins. Metchnikoff 's attempt to establish parallelism between phagocytosis and resistance. Work of his pupils. Metchni- koff 's interpretations. Origin of bactericidal substances from leuco- cytes. ' ' Macrocytase ' ' and microcytase. Metchnikoff 's interpretation of the Pfeiffer phenomenon. Origin of alexin. Leucocytic bacteri- cidal substances. Their nature. Leucocytic ferments. Leuco-protease. Petterson's experiments. Leucocytic extract of Hiss. CHAPTER XIII. — PHAGOCYTOSIS, Continued. FACTORS DETERMINING PHAGO- CYTOSIS 311 Opsonins. Tropins. Metchnikoff 's conception of stimulins. Work of Denys and his pupils. Other early observations. Work of Wright. Conception of opsonins definitely advanced. Analysis of opsonic action. Normal and immune opsonins. Neuf eld's opinions. Bac- teriotropins. Structure of opsonins. Specific absorption of op- sonins. Heat stability of immune opsonins. Eelation to other anti- bodies. Eelation to alexin. Variations in leucocytes as a factor in opsonic measurements. Kesistance to opsonic action on the part of bacteria. Eelation to virulence. CHAPTER XIV. — PHAGOCYTOSIS, Continued. OPSONIC INDEX AND VACCINE THERAPY 328 Wright 's work on typhoid immunization. Development of technique for measuring phagocytic activity. The phagocytic index. Opsonic index. Dilution method. Simon and Lamar's method. Accuracy of opsonic index. Wright's work on the staphylococcus infections. Eelation of opsonic index to clinical conditions. Negative phase. Summation of negative phase. Summation of positive phase. Clinical value of op- sonic index estimations. Opsonins and tuberculosis. Treatment by auto-inoculation. The value of opsonic index determinations. The xii CONTENTS PAGE value of vaccine therapy. Prophylaxis. Different types of infection and the logic of vaccine therapy in each type. The production and standardization of vaccines. CHAPTER XV. — ANAPHYLAXIS. FUNDAMENTAL FACTS 358 The relation of immunity to hypersusceptibility. Various kinds of hypersusceptibility. Historical development of our knowledge of these phenomena. The work of Eichet and others. The phenomenon of Arthus. The phenomenon of Theobald Smith. Experimental produc- tion of the anaphylactic state. Laws governing the condition as at first determined. Symptoms of experimental anaphylaxis in guinea pigs. Autopsy findings and causes of death. Changes in blood pres- sure. Changes in temperature. Leucopenia. Diminution of comple- ment. Symptoms in rabbits and dogs. Anaphylactic antigen. Spe- cificity of anaphylactic reaction. Quantitative relations. Variations depending upon method of administration. Anti-anaphylactic state. Prevention of anaphylaxis by drugs. Passive sensitization. Condi- tions governing its accomplishment. Quantitative studies of Doerr and Buss. CHAPTER XVI. — ANAPHYLAXIS Continued. FURTHER DEVELOPMENT AND THEORETICAL CONSIDERATIONS ....... 385 Theory of Gay and Southard. Besredka's theory. Gradual develop- ment of the antigen-antibody conception. Quantitative work. Identity of sensitizing and toxic substances. Idea of sessile receptors. Ana- phylaxis and precipitins. The work of Vaughan. Diminution of alexin during anaphylactic shock. Toxic substances obtained by action of active serum. Friedberger 's ' ' anaphylatoxin. ' ' Obtained from pre- cipitates. Obtained from bacteria. Is the mechanism of anaphylaxis intravascular or cellular? Precipitins and albuminolysins. Writer's opinion. THE MECHANISM OF ANTI-ANAPHYLAXIS. Nature of ana- phylactic poison. Peptone shock. PHENOMENA CLOSELY BELATED TO ANAPHYLAXIS. Toxicity of normal serum. Toxin hypersusceptibility. CHAPTER XVII. — ANAPHYLAXIS Continued. BACTERIAL ANAPHYLAXIS AND ITS BEARING ON PROBLEMS OF INFECTIOUS DISEASE . . . 410 Early work on sensitization with bacterial protein. Technique for sensitizing with bacteria. Bevision of our ideas of ' ' endo-toxin. " Vaughan 's work on toxic protein split-products. Friedberger 's ana- phylatoxin. Methods of production. Quantitative proportions which must be observed. Time and temperature conditions. Bearing of this work upon our underst nding of infectious disease. Friedberger 's interpretation. Bacteria, toxaamia. Is the bacterial antigen the matrix for the poison? CHAPTER XVIII. — ANAPHYLAXIS Continued. THE CLINICAL SIGNIFICANCE OF ANAPHYLAXIS . . . . . . . . . . 426 Serum sickness. Accelerated reactions and immediate reactions. Meth- ods of avoiding anaphylaxis in antitoxin injections. Anaphylaxis and bacterial vaccines. Asthma and hay fever. Sensitiveness to contact with certain animals. Possible anaphylactic reason for eclampsia. Sympathetic ophthalmia. Diagnostic reactions. Tuberculin reaction. Luetin reaction. Discussion of tuberculin reaction. Experimental ana- phylaxis with tuberculin. Diagnostic use of anaphylaxis. CHAPTER XIX. — THERAPEUTIC IMMUNIZATION IN MAN. THERAPEUTIC USE OF DIPHTHERIA ANTITOXIN 446 Statistical results. Amounts to be injected. Amount of antitoxin normally present in the human blood serum. PRACTICAL CONSIDERA- TIONS CONNECTED WITH DIPHTHERIA ANTITOXIN PRODUCTION AND STANDARDIZATION. Toxin production. L0 and L+ doses. Methods of determination. Production of antitoxin. Standardization of antitoxin, CONTENTS xiii PAGE U. S. Hygienic Laboratory method. Chemical concentration of anti- toxic serum. ACTIVE IMMUNIZATION IN DIPHTHERIA WITH MIXTURES OF TOXIN AND ANTITOXIN. Behring's work. Use of the method. Results obtained. INTRACUTANEOUS METHOD OF DETERMINING TOXIN AND ANTI- TOXIN VALUES. Principles of the method. Uses. Application of the method to the determination of antitoxin in human beings. TETANUS ANTITOXIN AND ITS STANDARDIZATION. Determination of the unit. ANTITOXIN AGAINST SNAKE POISON. Calmette's work. Differences between cobra and rattlesnake poison. Production of antiserum. PAS- SIVE IMMUNIZATION IN DISEASES CAUSED BY BACTERIA WHICH Do NOT FORM SOLUBLE TOXINS. General consideration of principles in- volved. Difficulties. Serum treatment of epidemic meningitis. Work of Kolle and Wasermann. Experiments of Jochmann. Flexner and Job- ling's experiments. Kesults. Present methods. Streptococcus anti- serum. Differences between various races of streptococci. Marmorek's serum. Work of Aronson, Tavel, Van de Velde and others. Probable manner of action. Serum treatment in pneumonia. Neuf eld's work. Eecent experiments and methods of Cole. Serum treatment of typhoid fever. Earlier experiments. Attempts to produce anti-endotoxin. Principles involved. Immunization with trypsin digested bacteria. Immunization with sensitized bacteria. Prospects of success. Serum treatment of plague. Yersin's attempts. Kolle and Martini's serum. Work of British Plague Commission. Lustig's serum. General results obtained. FACTS CONCERNING ACTIVE PROPHYLACTIC IMMUNIZATION IN MAN. General principles. Typhoid vaccination. Earlier history. Work of Wright, Kolle, and others. Russell's report of vaccination in the United States army. Statistics. Work of Metchnikoff and Bes- redka. Prophylactic immunization against cholera. Methods. Re- sults. Plague vaccination. Difficulties. Methods. Results. Small- pox vaccination. Rabies. Principles and methods of application. CHAPTER XX. — ABDERHALDEN 's WORK ON PROTECTIVE FERMENTS. MEIO- STAGMIN REACTION . 493 CHAPTER XXI. — COLLOIDS, by Professor Stewart W. Young, Stanford Uni- versity, California 499 Introduction. Definition. Reversible and irreversible colloids. Sta- bility of colloidal systems. Physical properties of colloids. Form and size. Osmotic pressure. Rate of settlement. Brownian movement. Electrical properties of colloids. Surface tension. Chemical properties of colloids. Flocculation of colloids by electrolytes. Salts and acid electrolytes. Influence of concentration. Diffe ence in sensitiveness to electrolytes. Explanation of phenomenon. T] e - " zone-phenomenon. " Mutual reactions of colloids. Mutual floccula ion. Protective action. Theories of interaction. The preparation of colloid solutions. Applica- tions to biology. Living tissues as colloids. Agglutination of bacteria. Analogy to colloid phhenomenon. Electrical charge carried by bacteria. Sensitiveness to light. Danysz phenomena. Conclusions. INFECTION AND RESISTANCE CHAPTER I INFECTION AND THE PROBLEM OF VIRULENCE THE early history of our knowledge of infectious disease is that of fermentation. It was a philosopher, Robert Boyle, writing in the 17th century, who prophesied that the problem of infectious dis- ease would be solved by him who elucidated the nature of fermenta- tion. His prediction was fulfilled 200 years later by the train of investigations begun by Cagniard-Latour and by Schwann, and car- ried to a brilliant culmination by Pasteur. It was the discovery of the living nature of ferments and the specific nature of the various micro-organisms which caused the several forms of fermentation, and especially of putrefaction, which made possible rational investi- gations in the field of infectious disease and led by analogy, first to logical speculation — then to actual experimental proof of the etiolog- ical relationship between the minute forms of life and the com- municable diseases. It is not much more than 50 years since Pollender described the anthrax bacillus in the blood and spleens of animals dead of this disease. In this short period the large number of maladies of ani- mals and human beings caused by micro-organisms belonging both to the varieties spoken of as bacteria and to those classified as protozoa has necessitated the segregation of this branch of knowl- edge into a separate chapter. The period of etiological investigation is now approaching its maturity. The causative agents of most of the more common infec- tious diseases have been discovered, and the biology of many of the pathogenic micro-organisms has been thoroughly studied both in their artificial cultures and in the infected animal body. In spite of a considerable accumulation of facts, however, the science of immunity, that is, the study of the defensive powers of the living animal body against infection, is still in its infancy, and the practi- cal therapeutic successes based on this science are disappointingly out of proportion to the really large amount of detailed knowledge of cellular and serum reactions at our disposal. The study of putrefaction and of fermentation — though furnish- ing the basic analogy from which the first impulse was obtained — 1 : INFECTION AND RESISTANCE presented after all a problem infinitely more simple than that of the infection of living tissues with bacteria. For, given any organic material containing suitable nutritive constituent, with favorable environmental conditions of moisture and temperature, and spon- taneously or experimentally inoculated with germs of a proper species, and the phenomena which ensued were merely those of bac- terial growth, in which an active part was played by the bacteria only, the dead organic materials serving simply as a passive men- struum for these activities. During the earlier days of the development of bacteriology, therefore, when the attention of investigators was concentrated prij marily upon the discovery of the specific causal agents of various in- fectious diseases, it seemed that the simple bringing together of pathogenic germ and susceptible subject should suffice for the ac- complishment of an infection. We have learned, however, that the process is much more involved, and that, fortunately for the sur- vival of the higher animals and man, the conditions which determine infection are intimately dependent upon a variety of secondary modifying factors. Throughout nature bacteria are abundant, and the environment of man and animals, the outer integuments of skin and hair, and the mucous membranes of the conjunctiva, the intestinal and respira- tory tracts, are constantly inhabited by a thriving bacterial flora. The distribution of certain species in definite localities is often suffi- ciently constant to be regarded as a normal condition. Thus the Bacillus xerosis is a characteristic inhabitant of the conjunctiva, certain cocci and spirilla are always present in the mouth and pharynx, as is Doderlein's bacillus in the vagina. The fact that bacilli of the colon group are invariably present in the bowels of ani- mals and man from the first few days or hours after birth has even been interpreted by some investigators as a physiologically beneficial condition. In the course of ordinary existence, therefore, and much more so during the course of accidental exposure to individuals in whom infection is present, the bodies of the higher animals are in intimate contact, not only with ordinarily harmless bacteria (sapro- phytes), but also with many varieties of the micro-organisms spoken of as "pathogenic" or disease-producing. Perfectly normal indi- viduals have, then, on occasion, been found to harbor diphtheria bacilli in nose and pharynx, meningococci have been found in simi- lar localities, and tetanus bacilli, the bacillus of malignant edema, the Welch bacillus, and other distinctly pathogenic germs have been isolated from the intestinal contents of individuals who showed no evidence of disease. In fact, the problem of the so-called bacillus car- riers — persons who, though themselves apparently well for the time being, harbor within their bodies and distribute to their environ- ment bacteria capable of causing disease in others — is, as we shall THE PROBLEM OF VIRULENCE 3 see, now recognized as one of the most important difficulties of sani- tary prophylaxis. In the case of typhoid fever this is particularly true, for it is now well known that a perfectly healthy individual may harbor typhoid bacilli in the gall-bladder for years and constitute, through all this time, a constant focus of danger to the public health. The accomplishment of an infection, then, is not determined merely by the fact that a micro-organism of a pathogenic species finds lodgment in or upon the body of a susceptible individual, but it is further necessary that the invading germ shall be capable of maintaining itself, multiplying and functionating within the new environment. An infection, then, or an infectious disease, is the product of the two factors, invading germ and invaded subject, each factor itself influenced by a number of secondary modifying circum- stances, and both influenced materially by such fortuitous conditions as the number or dose of the infecting bacteria, their path of en- trance into the body, and the environmental conditions under which the struggle is maintained. We have in truth, then, a battle of two opposed forces, the result of which is infectious disease. And it is the systematic analysis of these forces in their variable conditions, and the laws which govern them, which constitutes the science of immunity. It is the initial skirmish between the two which determines whether or not a foot- hold shall be gained upon the body of the subject and an infection thus established, and it is the balance between them which decides the eventual outcome of recovery or death. And though it is un- fortunately true that much of the knowledge gained by such studies has yielded no direct therapeutic results, the facts that have been revealed are fundamental to the pathology of infectious disease and as essential to the clinical understanding of these maladies as is the knowledge of the mechanism of the circulation, the chemistry of metabolism, or the structural changes of the tissues to the compre- hension of other pathological conditions. And from this point of view the study of infectious diseases can be made an eminently logical one, in that, knowing the criteria which govern the infection of a human being with a given germ, knowing the probable path of entrance, manner of distribution, and biological activities of the micro-organism, and the peculiarities of the mechanism of resistance set in motion in the body by this par- ticular infection, definite clinical deductions can often be made. One of the most fundamental facts, immediately apparent on considering the problems of infection, is the phenomenon that among the innumerable varieties of bacteria and protozoa present in nature there is a very limited group which is capable of becoming parasitic upon the body of higher animals, and among these a still smaller proportion which is capable of being "pathogenic" or causing dis- ease. We have used the terms pathogenic and non-pathogenic as 4 INFECTION ANE RESISTANCE practically synonymous respectively with "parasitic" and "sapro- phytic." But, as we shall see, although as a rule a micro-organism must be parasitic to possess pathogenic powers, some of the true saprophytes or so-called half-saprophytes may be pathogenic under certain conditions, and the terms do not cover each other absolutely. It is reasonable to suppose that all micro-organisms were origi- nally in the condition which we designate by the term "saprophytic." By this term we imply that these germs maintain themselves only upon dead organic matter and do not thrive in or upon the living animal tissues. The class of saprophytes is widely distributed and constitutes, of course, the most important group of bacteria in na- ture, since upon the activities of these germs depends the unlocking of nitrogen and carbon from the organic complexes in the dead bod- ies and waste products of animals and plants. Such bacteria if strictly saprophytic, that is, entirely unable to maintain themselves upon living tissues, have little importance as producers of disease, or, expressed in technical terms, have little "pathogenicity." Nev- ertheless, there are cases in which strict saprophytes may cause dis- ease by lodging upon and growing in animal tissues which have been killed by other causes, so-called necrotic areas ; and these, still being in relation with the body as a whole through the blood and lymph channels, furnish an area of saprophytic growth from which products of putrefaction or even bacterial poisons may be absorbed. While, as a rule, the disease following the invasion of necrotic tissue — such as gangrenous amputation stumps, old unhealed sinuses, diabetically gangrenous areas, etc., may be caused by a large variety of saprophytic bacteria, there are a few very important and specifically pathogenic bacteria which are, strictly speaking, saprophytes. Thus the form of meat poisoning caused by the Bacillus botulinus is due entirely to the poison formed by this bacillus outside of the body within the substance of the dead foodstuff, and disease ensues as the result of subsequent ingestion of this poison with the food. In the same way the tetanus bacillus and, less strictly speaking, the diphtheria bacillus, at least in its ordinary mode of attack, are rather closer to the class of saprophytes than to that of the parasites, since neither of these bacteria, under usual circumstances, invades the substance of the tissues beyond the point of initial lodgment, causing disease only by the production of specific poisons, a condition known as "toxemia" or intoxication. The tetanus bacillus, moreover, is not usually capable of maintain- ing itself and multiplying even at the point of initial lodgment unless the tissues have been injured by trauma or irritated by the presence of foreign bodies. Bacteria of such characteristics, there- fore, though pathogenic — that is, incitant of disease — remain never- theless essentially saprophytes living upon the dead animal tissues, not invading the living cells or body fluids. It is true that invest!- THE PROBLEM OF f^RULENCE 5 gations of Frosch 1 have shown that diphtheria bacilli may often be found in blood and organs of diphtheritic patients, and tetanus bacilli have occasionally been found in the spleen. However, such distribution is not necessary for the production of disease by these bacteria, and the essential point remains that they may cause violent, often fatal, disease without truly departing from their saprophytic mode of life upon dead tissues. Between the saprophytes and the true parasites or invaders of living tissue many transitions occur, and the condition of parasitism is probably a form of specific adap- tation. How such transition may be biologically developed is probably well illustrated by the investigations of Italian bacteriologists upon tetanus bacilli.2 Tarozzi 3 inoculated guinea pigs and rabbits with tetanus spores subcutaneously and found that these spores were rapidly transported to the liver, spleen, and kidneys, where they could maintain a latent existence for as long as 51 days. If during this period trauma or any injury of the organs was practiced which led to the formation of necrotic tissue the spores would develop upon this basis and cause acute or chronic tetanus. Canfora,4 continuing these studies, likewise found that tetanus spores inoculated under the skin are rapidly distributed throughout the circulation. If no trauma has taken place at the point of inoculation the locally lodged spores may be rapidly destroyed, probably by phagocytosis. In the circulation they appear to be less rapidly eliminated and may be present for from ten to thirteen days. If, during this period, there is produced a small wound, blood clot, or necrotic area in the body — this may serve as a focus for development and tetanus may ensue. After ten or more days the spores disappear from the blood, but may then take up a latent existence in some of the organs — as stated by Tarozzi. Apart from their importance as constituting a sort of transitional condition between pure saprophytism and parasitism, these investigations would seem to have much bearing upon the so- called cases of "cryptogenic tetanus." True infection, that is, the invasion of one species by individuals of another, and the ability of the latter to multiply and functionate within the cell complexes of the former, is a process quite out of keeping with the ordinary plans of nature, throughout which there seems to be a distinct opposition to the colonization and functiona- tion of one living being within the living substance of another. Thus, as Bail 5 has pointed out, a mass of frogs' eggs will remain 1 Frosch. Zeitschr. f. Hyg., Vol. 13, 1893. 2Belfanti, quoted from Canfora, Centralblt. f. Bact., I. Orig. Vol. 45, 1908. 3 Tarozzi. Centralblt. f. Bact., Orig. Vol. 38, 1905. 4 Canfora. Centralblt. f. Bact., Orig. Vol. 45, 1908. 5 Bail. "Das Problem der Bakt. Infection." Klinkhardt, Leipzig, 1911. 6 INFECTION AND RESISTANCE entirely uninvaded while alive, though the water surrounding it may swarm with bacteria of many varieties, but when by some accident such a mass of eggs ceases to live, it immediately falls prey to bac- terial infection. The same point is illustrated by the rapidity with which intestinal bacteria will spread throughout the body after death, when during life they have remained confined to the lumen of the intestine, or, at most, get into the portal circulation, to be de- stroyed in the liver. By the living cell, therefore, an opposition is offered to invasion by bacteria, a vital function which Bail has at- tempted to make clearer by formulating it as a law, referring to it as "Das Gesetz der Lebensundurchdringlichkeit." Upon what cell func- tion this vital resistance to invasion depends is to a large extent a mystery. It would seem to rest in principle upon the fact that the invading cell meets the invaded one under conditions peculiarly adapted to the activities of the latter, and is overcome before condi- tions suitable for its own activities have been established. The con- ditions here are not unlike those observed in the case of digestive enzymes, a comparison which becomes more than an illustrative anal- ogy when we consider that apart from the mere mechanical disturb- ance created by the presence of bacteria as foreign bodies the struggle between invader and tissue is largely one of enzyme against enzyme. Thus, for instance, the gastric juice does not act upon the mucous membrane of the stomach during life — but after death, at autopsy, partial digestion of this membrane by the pepsin is often seen. Whenever this vital resistance or opposition is overcome, and micro-organisms enter the tissues or cells, an abnormal process is taking place, and this process is, strictly defined, infection. Never- theless, it is by no means necessary that such infection should al- ways be accompanied by manifestations of disease. It is true that, in most cases, the natural resistance is such that a struggle ensues by which the invader is destroyed or thrown off, or in which the invaded subject is functionally injured or even killed, and the ac- companying evidences of such a struggle constitute what we know as infectious disease. But there are special cases, cases of adapta- tion, biologically speaking, in which neither invader nor host is seri- ously harmed.6 In the field of protozoology, especially, there are many examples of true parasites, that is, invaders truly maintaining their metabolism at the expense of the tissues and body substances of the host, which do not arouse reactions sufficiently vigorous to be termed "disease." Thus the Trypanosoma Lewisi may be found in the blood of rats 7 without noticeably affecting the health of the ani- mals, and other protozoa have similarly been found in organs and blood stream of a number of other apparently healthy animals. Al- though such conditions have been frequently spoken of as "infection 6 See also Bail, loc. cit. 7 Doflein. "Die Protozoen als Krankheitserreger." THE PROBLEM OF VIRULENCE 7 without infectious disease/7 the distinction is probably one of degree only — there being some reaction on the part of the host even in the mildest cases, if only in the weakening by withdrawal of body sub- stance, which distinguishes the infected from the uninfected animal. In other cases there may even be advantage to the host, following the infection, to the detriment of the invading micro-organism, a phe- nomenon most clearly illustrated by the invasion of the root hairs of leguminous plants by the Xitrogen-fixing " root-tubercle7' bacilli, a condition in which, as Fischer says, the plant may be regarded as parasitic upon the bacteria. The actual harm resulting from the infection must, to a large extent, depend upon the degree of adaptation to the new conditions of life possible on the part both of the invader and of the host. If the invader can acquire resistance to the defensive properties of the host, and the latter can be similarly adapted to the harmful effects of the invader, a prolonged condition of infection might ensue, a sort of truce without manifestations of the disease. Although this is con- ceivable, such mutual adaptation is probably very rare in human disease. In cases of so-called chronic septicemia in which bacteria may be again and again isolated by blood culture from the circulation it is more than likely that the organisms are constantly present, not because they multiply or maintain themselves within the circulation, but rather because they are being continuously discharged into the blood from an established focus in the tissues — as, for instance, on a heart valve. We have examined the serum of patients with subacute and chronic septicemia (endocarditis), and often found powerful opsonic action against the invading germs even when the patient's own serum and leukocytes were used in the tests, evidence that the bacteria were probably being successfully disposed of after they had gained entrance into the blood stream. In rabbits, too, in our ex- perience and in that of Miss Gilbert of this laboratory, it would seem that protracted septicemia is present only when secondary foci have been established from which the bacteria are constantly being dischargeiLinlQ_tha .blood. This we believe is rather the rule and the establishment of a balance within the blood stream an exception. When bacteria do succeed in withstanding successfully the opposing forces active within the circulating blood their rapid accumulation, the collapse of the defensive mechanism, and death of the patient are probably the most common course. The point of view which we have expressed in the preceding paragraph has been impressed upon us with particular insistence by the observation of certain cases of bacteriemia following infections of the middle ear, mastoid processes, and thromboses of adjacent veins. In such cases it appears that the blood may be flooded with bacteria which, nevertheless, disappear after the focus of infection has been 8 INFECTION AND RESISTANCE « removed. We have recently had the opportunity to observe this in a case of septicemia caused by Streptococcus mucosus, in which blood culture plates showed very numerous colonies, and in which recovery followed promptly upon complete excision of the thrombosed veins. It would seem to us, therefore, that bacteriemia offers a rather better prognosis than was formerly supposed, at least in cases in which the focus is surgically accessible. The same principle is illustrated in the ordinary clinical course of typhoid fever in the human being. Here the disease begins as a bacteriemia. Very rapidly, usually within two weeks, the bacteria disappear from the blood stream and a high serum immunity is established in the patient. Nevertheless, the bacteria remain actively growing within definite foci in the tissues, where they are to a cer- tain extent protected or inaccessible to the defensive powers so suc- cessfully active in the blood stream. At any rate the patient remains diseased and the bacteria can be isolated from the spleen, gall-blad- der, and intestines at a stage when they are no longer present in the blood stream, and during which a measurement of the bactericidal and opsonic powers of the patient will reveal a serum immunity much higher than normal. Just why the organisms are protected from these influences in the tissues we do not know. On the other hand, it is nevertheless true that a certain amount of actual adaptation between the bacteria and the body may take place and contribute to the chronicity of an infection. This seems to be shown especially by the experiments of Walker and others, which are referred to in other places, in which it was found that bac- teria grown on immune sera gain a certain amount of resistance against the injurious properties of these substances, and evidence more directly bearing upon the question is furnished by the studies on the typhoid carrier state in rabbits made by Chirolanza,8 Black- stein,9 Johnston, 10 and recently by Gay and Claypole.11 The last- named writers found that they could regularly produce the typhoid carrier state in these animals if they first cultivated the typhoid bacilli upon a medium containing defibrinated rabbits7 blood. Even in these cases, however, it is not at all improbable that the typhoid bacilli establish a permanent focus from which they are discharged into the blood stream. An infectious disease, therefore, may be interpreted as the result of parasitism in which no such mutual adaptation has taken place, and in which the invasion of the host by the micro-organism is marked by a struggle, the local and systemic manifestations of which constitute the disease. The disease is an evidence of conflict be- 8 Chirolanza. Ztschr. f. Hyg., Vol. 62, 1909. 9 Blackstein. Bull. Johns Hop. Hosp., 1891. 10 Johnston. Journ. Med. Ees., 27, 1912. 11 Gay and Claypole. Arch, of Int. Med., 12, 1913. THE PROBLEM OF VIRULENCE 9 tween the two forces, mild and locally limited if the protective pow- ers far outweigh the invasive powers of the micro-organisms, violent if the balance is reversed. This conception is probably a correct one in the case of the large majority of diseases — those in which invasion is accompanied by more or less rapid and violent inflammatory and other reactions. In diseases like leprosy, tuberculosis, and a few others of the more chronic infections it is also possible that extensive invasion of the body depends, not so much upon the active invasive ; powers of the micro-organism, powers which we will attempt to analyze presently, but rather upon the fact that for reasons of in- solubility and lack of irritating properties on the part of the invader! » no reaction is set up at first, and the invasion, though progressive, elicits no violent symptoms and no energetic opposition. The in- vader therefore progresses unopposed, becoming an incitant of dis- turbed bodily functions to the degree of actual disease only when it has gained a foothold in some organ and begun to proliferate, or has multiplied in such numbers that the cumulative effect of its toxic powers becomes manifest. Such a conception would assign the slow and gradual but pro- gressively invasive powers of such diseases as tuberculosis, leprosy, and syphilis in which systemic symptoms are manifest only after the disease has gained an extensive foothold, to the lack of acute physiological reaction resulting from the presence of the invading micro-organism. In the case of such infections as those caused by some of the yeasts or blastomyces we have seen foci of blastomycotic lodgment in the kidney and other organs surrounding which there was neither an accumulation of mobile cells — (leukocytes or lympho- cytes)— nor any evidence of cloudy swelling or other injury, by poisons, of adjacent parenchyma cells. Here, as in tuberculosis or leprosy, the reaction induced by the presence of the micro-organisms is slow and gradual — expressed in an eventual fixed tissue-cell reac- tion and giant-cell formation — similar to that induced by insoluble foreign bodies. And it may well be that the progressive ability to multiply without arousing the invaded body to rapid and powerful reaction may account for the prolonged period of apparent well- being in the early stages of such infections and permit the invaders to pervade the body so extensively. This point of view has been, we believe, most clearly expressed by Theobald Smith.12 Bacteria may lack invasive or pathogenic properties and be, therefore, immediately destroyed after gaining entrance to the host. They may be powerfully invasive and because of lack of adaptation arouse a violent defensive reaction on the part of the host. "There is another type of parasite," Smith says, "which may dispense largely with both offensive and defensive processes. We can conceive of this type as exerting a metabolic activity approx- 12 Theobald Smith. Journ. of A. M. A., May, 1913, Vol. 60. 10 INFECTION AND RESISTANCE imating so closely to that of the host that the latter reacts but slightly and then only after a long period of stimulation." Into this class he places the syphilis spirochseta and, in a somewhat modified sense, the tubercle bacillus. We have seen, then, that a micro-organism may be pathogenic and still be saprophytic in its mode of life. In order that this can occur, however, it is necessary that it should possess the power of producing at the place of lodgment a poison or toxin which can be absorbed and cause disease. The condition which ensues is not, prop- erly speaking, an infection, but rather a "toxemia'' differing from the toxemias resulting from the ingestion of drugs or other poisons only in so far as the toxins are manufactured at some point of bac- terial lodgment within the body of the victim. Typical tetanus and diphtheria, for instance, can be produced as readily by ingestion of the bacteria-free culture filtrates as by inoculation with the bacteria themselves. And although these bacteria may, on occasion, become invasive and thereby satisfy the criteria of true infection, this is not necessary for their pathogenicity. In the large majority of bacterial diseases, however, it is neces- sary that the germs shall be capable of producing a true infection before they can become pathogenic, and it is our task therefore to attempt to analyze those bacterial attributes upon which the invasive power or virulence may be said to depend. In the realm of infectious micro-organisms a wide range of cul- tural variations is encountered which indicates that some of these germs have adapted themselves very closely to the specific environ- mental conditions found in the living animal body, while others can take up with ease and under the simplest cultural conditions a purely saprophytic existence. Many pathogenic micro-organisms have so far defied all attempts at cultivation in artificial media. These we cannot use for examples since it may well be that the failure of attempts in many of them may hinge upon such simple alterations of method as the exclusion of oxygen, the addition of fresh tissue, or the supplying of amino- acids, which have made possible the cultivation of the spirochseta pallida and the leprosy bacillus. But among those which we can cultivate there are many which require for successful cultivation the production of artificial conditions simulating closely those ob- taining in the living body. Thus malarial plasmodia can be made to multiply only if furnished with uninjured human red blood cells, within which they can develop. The gonococcus requires, in its first cultures outside the body, a medium containing human protein ; and the hemophile bacteria, among them the influenza bacillus, re- quire hemoglobin. Other organisms like pneumococci, many strep- tococci, diphtheria bacilli, and many others, though easily grown on artificial media, are still fastidious in their requirements and develop THE PROBLEM OF VIRULENCE 11 sparsely or not at all unless definite conditions of nutrient materials, temperature, reaction, and osmotic pressure are observed. On the other hand, typhoid, anthrax, and dysentery bacilli, staphylococci and numerous other pathogenic germs grow easily and luxuriantly on the simplest laboratory media and within a wide range of envi- ronmental variations. Biologically considered, we could arrange the scale of adaptation to parasitic conditions on this basis and it would seem, a priori, that those bacteria which had thus adapted themselves most closely to the living body should be the most infectious. There is not, however, such parallelism, since many of the most powerfully invasive or virulent germs, for instance, the anthrax bacillus, have retained their capacity for saprophytic life to the fullest extent. It is more logical, therefore, to classify parasites, not according to their ability to revert to saprophytic conditions, but rather, as Bail 13 has done it, on the basis of their relative powers of invading the living body. His classification, of course, implies that the position of each micro- organism in this scale must be determined with reference to a given animal species, since a germ which is highly infectious ("parasitic" in Bail's sense) for one species may be a "half-parasite" or even a pure saprophyte for another. Briefly reviewed, his classification is as follows: I. Pure Saprophytes. — (Xecroparasites, superficial parasites, or external parasites.) Micro-organisms which under no circumstances can be made to develop within the living tissues of a given animal. This does not exclude their pathogenicity for this animal, since, like the diphtheria or tetanus bacillus, they may develop and produce toxins on the basis of a localized area of dead tissues. II. Pure Parasites. — Organisms like the anthrax bacillus or the bacilli of the hemorrhagic septicemia group which, implanted in small quantity in an animal, will rapidly gain a foothold, thrive, and spread throughout the body. III. Half parasites, organisms which may be infectious if in- troduced into the animal body, but, not possessing this invasive power to the same degree as the preceding class, require the inocula- tion of considerable quantities, often a special mode or path of in- oculation, or even possibly a preliminary reduction of the local and general resistance of the infected individual in order that they may multiply and become generalized. This class includes the large majority of the bacteria pathogenic for man. This property of invasive power is spoken of as virulence in contradistinction to toxicity — the latter implying merely the abil- ity to produce poisons, and not necessarily being associated with the power to invade. 13 Bail. Loc. cit. IS INFECTION AND RESISTANCE In order that a micro-organism may be a true parasite in Bail's sense — or invasive — for any given species of animal it must of course possess certain basic cultural attributes which enable it to grow in the environment furnished by the host. For instance, a micro-organism which does not grow at temperatures below 37.5° C. cannot very well become parasitic upon cold-blooded animals. An excellent illustration of this influence of body temperature upon the invasive powers of bacteria is furnished by the different races of acid-fast bacilli which invade the bodies of man and of birds. The avian tubercle bacillus, for instance, is non-pathogenic for man and in cultures will not develop at temperatures below 40° C., which is about the body temperature of most birds. The human tubercle bacillus, on the other hand, is non-pathogenic for birds and ceases to grow in artificial cultures when the temperature is raised above 40° to 41° C. This is merely one of a number of examples which might be cited to demonstrate the necessity of simple cultural adaptation, as it influences the property of virulence. Again, it is probable that in order to develop in the animal body it is necessary that a micro-organism shall be capable of developing without free oxygen. While this point is not definitely certain, it is not probable that any of the virulent bacteria can be strict aerobes. As a matter of experience none of the pathogenic bacteria at present known are absolute aerobes — though many of them grow better in artificial cul- ture when oxygen is freely present than when it is absent. Furthermore, the conditions encountered by bacteria as they enter the animal body will vary considerably according to the path by which they gain entrance. Organisms entering by the intestinal canal are subjected to conditions of acidity or alkalinity, the action of digestive juices, of bile, and to competition with other intestinal bacteria, forces to which many pathogenic germs will succumb, while others may survive there and thrive. Those entering into the tissues by way of the skin and mucous membrane, on the other hand, en- counter an immediately mobilized protective mechanism which, suc- cessfully resisted by some of them, might easily and quickly dispose of small quantities of other bacteria more resistant to conditions in the bowel. It is but natural for this reason that the accomplishment of an infection by any given germ must depend to a great extent upon its gaining entrance to the body by the path best adapted to its peculiar requirements. The mechanical protection afforded by the coverings of skin and mucous membranes is as a rule sufficient to prevent the penetration of any bacteria which by chance may have found lodgment upon them. In the case of the most usual pyogenic cocci and many bacilli such protection is probably absolute, and a distinct break of con- tinuity, such as a bruise or a wound, even though this may be too small to attract attention, is necessary for successful infection. In THE PROBLEM OF VIRULENCE 13 the case of a very limited number of diseases infection seems to take place even through the unbroken skin, and the method, often spoken of as the vaccination method of Kolle, employed in many instances when it is desired to produce experimental plague infection in rats or guinea pigs, consists in merely rubbing a small amount of cul- tural material into a shaven area of the skin. However, in, this case, as well as in other instances where mere massage of bacteria into unbroken skin has led to successful inoculation, it is more than likely that success has depended upon either microscopic lesions or possibly the violent introduction of the organisms into the sebaceous glands, the sweat glands, or hair follicles. The defense of intact mucous membranes, however, is by no means impervious. While many organisms can be implanted upon mucous membranes with impunity, there are a number of others that can cause local inflammations upon these and can further pass through them into the deeper tissues and thence into the general system. Thus gonorrhea is ordinarily a disease of implantation upon a mucous membrane, and diphtheria bacilli and streptococci give rise to localized disease on the pharyngeal and nasal mucosse, the latter not infrequently penetrating from the initial point of lodgment upon the mucosa into the deeper tissues and the circulation, causing a condition of "septi- cemia" or abacteriemia." For the experimental determination of the penetrative power of organisms through mucous membranes the conjunctiva has been a favorite test object, and it has been shown that plague 14 and glanders,15 as well as hydrophobia, may be trans- mitted by simple instillation of infectious material into the unin- jured conjunctival sac. In the case of hydrophobia 16 it is related that in Paris a young man contracted hydrophobia by rubbing his eyes with a finger contaminated with the saliva of a rabid dog. In the case of syphilis, though often claimed, there is no positive proof to show that infection may take place through the uninjured sur- faces. It has been definitely shown, however, that tubercle bacilli 17 may pass into the lymphatics through the intestinal mucosa without there being any traceable injuries on this membrane. It may well be, however, that even without the existence of demonstrable morphological lesions penetrability by micro-organisms may presuppose local physiological or functional injury, such as con- gestion or catarrhal inflammation. Thus it is seen that the mechanical obstacle to the entrance of micro-organisms offered by skin and mucous membranes, though important and not to be underestimated, is by no means a perfect safeguard. 14 Germ. Plague Com. Arb. a. d. kais. Gesundheitsamte, Vol. 16, 1899. 15 Conte. Rev. veterin., Vol. 18, 1893. 16Galtier. Compt. rend, de la soc. biol., 1890. "Bartel. Wien. Klinikhandt, 1906-1907. 14 INFECTION AND RESISTANCE However, it is only very definite species of micro-organisms which can cause disease at all when introduced into the body by these paths. For, although the rubbing of plague bacilli into the skin, or the inoculation of a cut surface with streptococcal or glan- ders bacilli, will rapidly lead to progressive infection, similar inocu- lation with the typhoid bacillus or the cholera spirillum would lead to no such result. And, though the swallowing of pus cocci, pneu- mococci, and a number of other micro-organisms would be entirely without effect, similar ingestion of the typhoid and cholera organism would usually result in typical infection. The path of introduction, therefore, is an important considera- tion in determining whether or not a given micro-organism may give rise to disease. It is necessary that the manner of gaining entrance be suited to the cultural and other peculiarities of the germ in ques- tion. In the case of cholera, for instance, the spirillum which causes this disease is peculiarly susceptible to the deeper defences residing in the body fluids and cells, and cutaneous infection by the small numbers of bacteria likely to be introduced in this way would promptly be checked by these agencies. In the intestinal mucosa, however, the cholera spirillum finds conditions most favorable for rapid multiplication, and the disease is caused by the inflammation and destruction of the mucous and submucous tissues by the poison- ous substances emanating from the large numbers of cholera spirilla which die and are disintegrated, as well as by the absorption of these poisons into the circulation. The bacteria themselves, however, never gain a permanent foothold within the blood or other organs. In the case of typhoid fever the conditions are somewhat similar, although here, during the earlier weeks of the disease, we have an actual penetration of the bacilli into the circulation. This, however, prob- ably takes place only after intraintestinal proliferation has taken place, which then, on' the injured mucosa, represents a dose out of all proportion great when compared with the quantities that would spontaneously come into contact with the external surface of the body. This leads us to another important factor concerning the invad- ing forces, in the determination of successful infection, namely, that of the quantity introduced or the dosage. In order to cause infection, even when the bacteria are of the variety known to produce disease or "pathogenic," and are brought into contact with the body by a path suitable to their peculiar re- quirements, the initial quantity introduced must be sufficiently large to preclude complete annihilation by the first onslaught of the de- fensive powers of the body. It is plain, therefore, that in the case of bacteria weak in power to cause disease, given the subject of in- fection and his defences as a constant, the quantities to be introduced must be larger than in the case of micro-organisms of violent disease- THE PROBLEM OF VIRULENCE 15 producing properties. The dosage necessary to cause infection, therefore, is in inverse proportion to that property of bacteria spoken of as their "virulence." Thus we measure the degree of the so-called virulence of bacteria by determining the smallest quantity, measured by dilution of platinum loops or by fractions of agar slant cultures (both very inexact methods), which will still cause infection and death in susceptible animals of a standard weight. In the case of micro-organisms of extreme virulence, such as the anthrax bacillus or bacilli of the hemorrhagic septicemia group, the inoculation of a very small number of bacteria may suffice to initiate infection. In- deed, it has been claimed for the anthrax bacillus that the injection of a single bacterium will produce fatal disease in a susceptible animal. The inverse relation existing between the degree of viru- lence and the number of bacteria inoculated is well illustrated by the experiments of Webb, Williams, and Barber,18 carried out upon white mice with anthrax, by the method of inoculation devised by Barber.19 This technique consists in picking up single organisms with a capillary pipette under microscopic control, from a very thin emulsion of bacteria and injecting directly from the pipette through a needle puncture in the skin. While requiring a considerable de- gree of skill, the method, when successful, permits an actual accurate count of injected bacteria instead of the merely approximate esti- mate which can be made by consecutive dilutions of thicker emul- sions. In their experiments with anthrax in white mice Webb, Wil- liams, and Barber found that the inoculation of a single thread of anthrax bacilli (3 to 6 individuals) taken directly from the blood of a dead animal (that is, in the most virulent condition) would regularly cause death, and it was impossible for this reason to immunize with such bacilli. On the other hand, if taken from 12-hour agar cultures of the same strain such small quantities would often fail to kill. The brief period of growth under artificial conditions had sufficiently lessened the virulence of the bacilli so that 2, 3, and more threads could be injected with- out harm. And after several generations of such cultivation as many as 27 and more threads could be inoculated with im- punity. Another example of the measurement of relative degrees of- virulence, by a method more commonly employed, may be illustrated as follows : The problem in which this particular measurement was used consisted in the comparison of the virulence of two strains of pneumococcus, one (N2) successively passed through white mice, the other (N\) kept alive for several weeks on serum-agar. To accom- plish this graded quantities of 18-hour broth cultures of the two 18 Webb, Williams, and Barber. Jour. Med. Res., 1909, Vol. XV. 10 Barber. Kansas Univ. Science Bulletin, March, 1907. 16 INFECTION AND RESISTANCE strains were injected into mice of approximately the same weight as follows:20 Ni Result N2 Result 0.1 c. c. = dead 24 hrs. 0.1 c. c. = dead 24 hrs. 0.05 c. c. = lives 0.05 c. c. = dead 24 hrs. 0.02c. c. = lives 0.02 c. c. = dead 24 hrs. 0.01 c. c. = lives 0.01 c. c. = lives This example further illustrates another important fact in con- nection with the problem of infection — namely, that within the same species of bacteria different races of strains may exhibit widely varying degrees of virulence. This has been known since the days of Pasteur, and it is indeed of great importance in the immunization of animals that weakly virulent strains of a given micro-organism may be used to produce a gradual immunity against the same species of bacteria in their fully virulent condition. Though observed in almost all species of bacteria such variations are especially notice- able in the cases of streptococci and pneumococci — organisms in which no two strains may be alike in infectiousness, and in which the injection of some strains into susceptible animals may produce no result whatever, while other strains will kill if administered in the smallest measurable quantities. To a large extent these fluctua- tions of virulence appear to represent degrees of adaptation on the part of the bacteria to the conditions met with in the living body; and the ease with which such variations can often be artificially pro- duced would seem to furnish another proof that the property of in- fectiousness is a biological attribute of relatively recent acquisition. For, although no general statement of absolute accuracy can be made, it is a fairly uniform rule that races of pathogenic bacteria gain in virulence as they are passed through successive animals of the same species, and lose in virulence as they are preserved upon media under conditions of artificial cultivation. Further showing this ability to rapidly adapt themselves is the observation that passage through animals of a certain species will enhance the virulence for this species, but often reduce it for animals of another kind. Among the earliest observations on this point are those of Pasteur 21 in his work on rabies. He found that the virus of hydrophobia when successively passed through rabbits gained in virulence until a degree of maximum infectiousness was attained 20 For making such accurate measurements we have recently found very useful the Precision syringe described by Terry, Jour, of Inf. Dis., Vol. 13, 1913. 21 Pasteur and Thuillier, Compt. rend, de I'acad. des. sc., Vol. XII, 1883. THE PROBLEM OF VIRULENCE 17 beyond which it could no longer be enhanced. After only three passages through monkeys, however, the virulence of this "virus fixe" for rabbits was reduced almost to extinction. His experience with swine plague was similar. Swine plague bacilli successively passed through rabbits and pigeons gained enormously in virulence for these animals respectively, but lost in virulence for hogs. There are numerous methods by which the virulence of micro- organisms can be attenuated by laboratory manipulations, and since many of them are of great importance in the active immunization of animals we will reserve their detailed discussion until we come to consider the methods of immunization themselves. Suffice it to say in this place that most methods of attenuation consist in subjecting the bacteria, in artificial culture, to deleterious influences, either of unfavorably high temperature, exposure to light or harmful chemi- cal agents, or allowing them to remain in prolonged contact with the products of their own metabolism by infrequent transplantation. As a rule the attenuation which inevitably follows any form of arti- ficial cultivation in the case of bacteria like streptococci or pneumo- cocci can be delayed by preserving them in media containing sera or tissues. In the case of the pneumococcus, for instance, one of the best methods of conserving virulence in storage is to keep them either in a soft rabbit-serum-agar mixture, as practiced by Wads- worth, or, better still, to store them within the spleen of a mouse dead of pneumococcus infection, as recommended by Neufeld. The mouse is autopsied and the spleen kept in the dark and cold in a des- iccator, under sterile precautions. This, again, as well as the en- hancement of virulence on passage through the same species of ani- mal— or the reduction of virulence for one species by passage through another — shows that such fluctuations are dependent upon a very delicate biological adaptation. It is interesting, moreover, to look upon this process of adapta- tion as a sort of immunization of the bacteria against the defensive powers of the host, a conception early suggested by Welch. For just as the animal body may become more resistant to the offensive weapons of the invaders, so it is reasonable to suppose that the bac- terial body may gradually develop increased resistance to the de- fensive mechanism of the host. And this, if it occurs, would of course lead to an increase of its invasive power or virulence. The increase of virulence by passage through animals would alone lead us to suspect that such acquired resistance to destructive agents on the part of the bacteria might be responsible for the enhancement, but additional evidence pointing in this direction has been brought by experiments in which it was shown that bacteria cultivated in the serum of immune animals not only gained in resistance to destruction by the serum constituents, but at the same time were rendered more highly pathogenic. Experiments of this kind were carried out by 18 INFECTION AND RESISTANCE Sawtchenko,22 by Danysz,23 and by Walker.24 The results of Wal- ker are especially instructive. He worked with a typhoid bacillus which he cultivated for a number of generations upon the serum of a typhoid-immune animal, and found that after such treatment the organism had gained in virulence and lost in aggiutinability by im- mune serum, and that a larger amount of specific immune serum was necessary to protect animals against it than sufficed for protection against normal typhoid strains not thus cultivated. We will refer to these results more In detail in a later chapter, since the conception will be easier to grasp when we have considered more fully the mechanism of defence at the disposal of the animal body. That this power of gaining resistance against deleterious influ- ences on the part of bacteria is not confined to their resistance to the animal defences alone is well shown by the experiments of Danysz 25 upon the immunization of anthrax bacilli against arsenic. In in- oculating series of 50 tubes containing arsenic dilutions (ranging from 1 to 10,000 to 1 to 200) with anthrax bacilli Danysz found that up to 1 to 5,000 the arsenic increased the growth of the bacilli ; in concentrations higher than this growth was inhibited. By grad- ually progressive cultivation of the organisms in increasing concen- trations of arsenic he finally succeeded in obtaining growth in solu- tions five times more concentrated than those in which they would develop at first. It is intensely interesting also that Danysz found, both in the case of his serum-resistant and arsenic-resistant strains, that, as they became less sensitive to the deleterious effects of these agencies, they were altered morphologically in that they developed capsules. Similar in significance to this is the very important observation that certain strains of spirochseta pallida may acquire resistance against salvarsan or "606." 26 These so-called arsenic-fast strains are ap- parently unaffected by the injection of this preparation into the patient. The experiments of Danysz were probably the first to call atten- tion to the possible relationship of bacterial capsule formation to virulence, and this particular phase of the subject has since then been extensively studied. It is a matter of common observation that micro-organisms like the pneumococcus, the anthrax bacillus, some streptococci, and a number of other germs which are capable of pro- ducing capsules under suitable conditions are most virulent in the capsulated stage. As the strains are passed through animals and their virulence increases their ability to form capsules becomes more 22 Sawtchenko. Ann. Past., Vol. 11, 1897. 23 Danysz. Ann. Past., 14, 1900. 24 Walker. Jour, of Path, and Bact., Vol. 8, 1903. 25 Danysz. Loc. cit. 26 Oppenheim. Wien. kl. Woch., 23, 1910, No. 37. THE PROBLEM OF VIRULENCE 19 and more apparent — whereas the diminution of virulence which takes place on artificial media is accompanied by a gradual loss of capsule formation. Organisms like the Friedlander bacillus which retain their ability to form capsules almost indefinitely in artificial culture moreover do not lose their virulence to any great extent as long as this property is preserved. It is also well known that cap- sulated bacteria are peculiarly insusceptible to the ordinary aggluti- nating powers of specific immune sera. This has been noticed, not only in the case of heavily capsulated bacteria like those of the Fried- lander group or the streptococcus mucosus, but, in the case of plague bacilli — where capsulation is usually present only in cultures taken directly from the animal body and cultivated at 37° C., Shiba- yama 27 has found a direct relation between non-agglutinability and a slimy condition of the cultures. Cultures kept at 5° to 8° C. in the ice-chest were easily agglutinable and lacked the slimy property. Cultures kept at 37.5° C. were slimy and thready in consistency and were not as easily agglutinated by the same immune serum. Forges 28 later showed that inagglutinable, capsulated bacteria can be made amenable to the agglutinating action of the serum — which we may assume to indicate vulnerability by the serum if the capsule is previously destroyed by heating at 80° C. for about 15 minutes in % normal acid. Against the cellular defences, the leukocytes, capsulated bacteria seem also to be more resistant than are the non-capsulated. This has been especially studied by Gruber and Futaki,29 who find that a capsulated bacillus is rarely taken up by a phagocyte even when these cells are apparently normal and able to take up the uncap- sulated organisms. They go so far as to claim that, in the case of anthrax in rabbits, the development or absence of a capsule deter- mines whether or not infection can take place. The same conclusion is reached in similar studies by Freisz,30 who does not believe that anthrax bacilli can ever cause infection unless they possess the power of forming capsules. All this experimental evidence points strongly toward a probable direct relationship between capsule for- mation and virulence, in the sense that a thickening of the ectoplasm may in some way protect the bacteria from the destructive forces aimed at them by the cells and fluids of the invaded body. As a matter of fact, even when no distinct capsule is visible, it is nevertheless possible that ectoplasmic changes may take place. This phase of the subject has been thoroughly discussed by a number of writers, more especially by Eisenberg.31 It appears that many 27 Shibayama. Centralbl f. Bact., Orig. Vols. 38, 1905, and 42, 1906. 28 Forges. Wien. klin. Woch., p. 691, 1905. 29 Gruber and Futaki. Munch, med. Woch., 6, 1906. 30 Preisz. Centralbl. f. Bakt., Vol. 49, 1909. 31 Eisenberg. Centralbl. f. Bakt., I, 45, 1908, p. 638. 20 INFECTION AND RESISTANCE bacteria, in which true capsule formation has not been observed, may show swelling or enlargement under conditions in which their of- fensive activities in the infected animal body are called into play.32 Radziewsky33 has noticed such swelling of B. coli in fatal guinea- pig infections, and spoken of it as "one of the characteristic signs of infectiousness." Kisskalt 34 has described the same thing in the case of streptococci, and Eisenberg interprets this as signifying an ectoplasmic hypertrophy comparable in principle to capsule forma- tion. He looks upon the ectoplasmic zone as a protective layer, and calls attention to the observation of Liesenberg and Zopf)35 who showed that capsulated strains of leukonostoc mesenteroides will withstand 85° C., a temperature at which uncapsulated forms are rapidly killed. There is a considerable amount of evidence, then, which seems to indicate that the development of a capsule is at least one im- portant method by which the bacteria can protect themselves against the onslaught of the defences of the invaded animal body and in so doing become more virulent. It is not likely, however, that this merely passive increase of the resistance to injury on the part of the bacteria accounts for the entire train of phenomena included in an enhancement of virulence. It has been suggested by a number of observers that definite active offensive characteristics distinguish the virulent from the avirulent bacteria, in that the former may secrete, within the living body, substances by which the destructive powers of serum and leukocytes are neutralized or held at bay. A very definite suggestion of such a possibility we find expressed in the now classical paper of Salmon and Smith36 on hog cholera immunity, published in 1886. They say : ". . . the germs of such maladies are only able to multiply in the body of the individual attacked, because of a poisonous principle or substance which is produced during the multiplication of these germs." 37 Bouchard formulated such a theory in 1893 by speaking of the "produits secretes par les microbes pathogeniques," substances which he found in cultures of virulent bacteria, and which seemed to reenf orce the invasive powers of the germs. Kruse 38 also within the same year developed a similar idea. He assumed that bacteria may secrete enzyme-like substances which paralyze the destructive properties of animal serum, and in this way gain the power to i2 These forms Bail has spoken of as. "thierische Bazillen." 33 Radziewsky. Zeitschr. f. Hyg., Vol. 34. 34 Kisskalt. Cited after Eisenberg, loc. cit. 35 Liesenberg and Zopf. CentralbL f. Bakt., Vol. XII, 1892. 36 Salmon and Smith. Proc. Biol Soc., Washington, D. C., Ill, 1884, 6, p. 29. 37 A typewritten copy of this paper was kindly put at my disposal by Prof. Theobald Smith. 38 Kruse. Ziegler>s Beitrage, Vol. XII, 1893. THE PROBLEM OF VIRULENCE 21 invade. As a matter of fact we have learned, since that time, that staphylococci may secrete soluble substances, "leukocidins," which injure white blood cells, and that many bacteria produce similar poisons, "hsemotoxins," which specifically injure red blood cells — thereby causing anaemia and reducing the resistance of the host. However, the correlation and further elaboration of these thoughts of Salmon and Smith, of Bouchard and of Krnse was left to Bail,39 in what is known as his "aggressin theory." Bail maintains on the basis of careful experimentation that virulent bacteria can produce within the animal body substances which he calls "aggressins," upon which depend their invasive powers or virulence. These substances are secreted only under stress of the struggle against the unusual defences, are not demonstrable in test-tube cultures, and are in themselves, according to Bail, entirely non-toxic. He obtains these aggressins by injecting virulent bacteria into the peritoneal cavity of a guinea pig and immediately after death re- moving the exudate. This he centrifugalizes, removes the bacteria and cells, and sterilizes the supernatant liquid by the addition of small quantities of chloroform. The action of the exudates in which aggressins have been produced by the bacteria is the following: (We take this tabulation from Bail's own paper on typhoid and cholera aggressins in the Archiv fur Hygiene, \7ol. 52, p. 342.) 1. Sublethal doses of typhoid bacilli or cholera spirilla become lethal when the aggressin is injected with them. 2. Lethal doses of bacilli which ordinarily would cause a slow infection only cause a rapid and severe' infection when aggressins are added. 3. The addition of aggressin neutralizes the bacteria-destroy- ing power of immune serum in the peritoneal cavity of a guinea pig. 4. The injection of aggressin alone produces subsequent im- munity. It is impossible to discuss with completeness the arguments ad- vanced for and against the correctness of Bail's views until we have described in detail the mechanism of protection at the disposal of animals. But the main objection brought against this theory is that of Wassermann and Citron,40 who claim that all these properties of the aggressive exudates can be explained by the fact that they con- tain extracts of the bacteria (endotoxins), which, injected wTith a sublethal dose of bacteria, merely enhance their action in the same way that this would have been accomplished by the injection of additional dead bacterial bodies. It will require much further work before this point is settled, and the problem is peculiarly involved and 39 Bail. Archiv f. Hyg., Vols. 52 and 53, 1905; Folio serologica, Vol. 7, 1911. 40 Wassermann and Citron. Deutsche med. Woch., Vol. 31, 28, 1905. 22 INFECTION AND RESISTANCE difficult. However, the recent work of Rosenow 41 on pneumonococci seems to bring some reinforcement to the ranks of those who main- tain the existence of a special offensive substance at the command of virulent bacteria. Rosenow extracted pneumococci grown on serum broth and found that such extracts when made from virulent strains would protect avirulent strains from engulfment by phagocytes. The non-virulent strains left in these extracts for 24 hours became viru- lent. He believes, therefore, that the virulence of pneumococci de- pends largely upon the possession of these substances which he calls "viruiins," and which in function at least are conceived as very similar to the "aggressins." Recent results obtained by the writer 42 with Dwyer seem to indicate that anaphylatoxins produced from the typhoid bacillus possess some of the properties claimed for his aggressin by Bail. It is not impossible that the "aggressins" obtained by him were of this nature. Virulence, then, may be analyzed into two main attributes : one a purely passive property of resistance or self-preservation on the part of the bacteria, perhaps morphologically expressed in ectoplas- mic hypertrophy and capsule formation; the other an actively of- fensive weapon in the form of substances of the nature of the "ag- gressins" of Bail or the "virulins" of Rosenow. The extent of our present knowledge of details does not warrant a statement of the case in more definite terms. From the facts we have discussed in the preceding paragraphs it now becomes manifest that the elements which determine the nature of an infectious disease are twofold. On the one hand each variety of infectious germs possesses certain biological and chemical attributes which are specific and peculiar to itself ; by these its predilection for path of entrance and mode of attack is de- termined, and upon these depends the nature of the reaction called forth in the animal body. On the other hand the degree of infec- tion in each case, the severity of the reaction and the ultimate out- come are determined by the balance which is struck between the virulence of the entering germ and the protective mechanism opposed to it. The specific properties of each micro-organism are the factors which account for the clinical uniformity (within definite limits) which is observed in the maladies produced in different individuals by the same species of bacteria. Thus a severe typhoid fever is, in essential characteristics, entirely similar to a mild case — since in both instances the path of entrance, through the intestine, is the same, the distribution of the germs after entrance differs only in degree, and the reactions, local and systemic, which are called forth 41 Rosenow. Jour, of Inf. Dis., Vol. 4, 1907. 42 Zinsser and Dwyer. Proc. Soc. Exp. Biol. and Med., Feb., 1914. THE PROBLEM OF VIRULENCE 23 are alike. And cases of this disease in general differ as a class from the maladies caused by, let us say, the group of clinical conditions resulting from anthrax infection, where entrance is through the skin, and generalized infection of the blood ensues without definite or regular localization in any given organ. Again, a localized staphylococcus abscess will differ materially from an equally local- ized focus of tuberculosis, because the chemical constituents of these bacteria respectively call forth each a characteristic response on the part of the defensive mechanism. Such specificity of the various micro-organisms may of course be due partly to their mode of attack and distribution, and partly, as we shall see, to the pharmacological action of the poisonous prod- ucts given out by them. That both factors contribute seems beyond doubt; but recent work, especially that of Friedberger, which is fully discussed in another place (see p. 413), seems to show that clinical differences depend much less than was formerly supposed upon specificity of the intracellular poisons, and much more upon distribution and localized accumulation of the germs, conditions which are determined rather by the mode and extent of invasion than by chemical differences of poison production. This problem, rather difficult to discus^^on the limited basis of the facts so far outlined, will become clearer 'as we proceed, but we need only refer at present to the essential clinical uniformity of the various forms of septicemia, where organisms freely circulate in the blood — with often a focus of distribution on a heart valve — conditions in which it is rarely possible to determine the species of the responsible germ except by blood culture. Or, again, as Friedberger 43 points out, there is great similarity between the ordinary pneumococcus pneumonia and that caused by the Fried- lander bacillus. In both cases the distribution and mode of attack of the bacteria are essentially the same, though the micro-organisms' themselves are biologically very dissimilar. One and the same micro-organism, on the other hand, may cause entirely different clinical conditions, and here the type of infection depends purely on the degree of invasion possible in the given case — that is, the balance between virulence and resistance. A germ may enter the body and cause an inflammatory reaction at the point of entrance, the process remaining purely localized. In such cases the defensive forces have been so efficient, the invasive properties of the germ so relatively weak, that progression beyond the point of en- trance is prevented and the resultant disease takes the form merely of a localized abscess. This is the case when a healthy individual is infected with an attenuated organism or by one whose species' characteristics do not include a powerful invasive property. Thus streptococci, if entering the tissues of a normal subject in small 43 Friedberger. Deutsche med. Woch., No. 11, 1911. 24 INFECTION AND RESISTANCE numbers or in attenuated form, may produce a purely localized in- fection, and ordinarily non-pathogenic germs like proteus, subtilis, or colon bacilli may produce localized abscesses in weak and debili- tated individuals, though implanted upon a healthy subject they would be rapidly disposed of without gaining even a preliminary foothold. Such tendency to localization is the common form of in- fection in the case of a number of germs. It is the most usual type of staphylococcus infection, for instance, in which the degree of virulence of the strains ordinarily met is such that the balance struck by them with the average defensive powers of man results in localiza- tion. However, the same micro-organism, enhanced in virulence, or gaining entrance in unusual numbers in a weakened individual, may rapidly spread from the point of inoculation, at first by contiguity, then by invasion of the blood and lymph channels, and become generalized. When organisms become generalized and circulate in the blood; the resulting condition is spoken of as septicemia or bacteriemia. This is the form of infection commonly caused by streptococci, bacilli of the hemorrhagic septicemia group, anthrax bacilli, and many others. It implies a powerful invasive property and always constitutes a condition of great- gravity when persistent. We are learning of recent years, however, that in many infectious diseases formerly regarded as purely localized a temporary entrance of the bacteria into the circulation is a usual occurrence. Thus Fraenkel 44 has shown that lobar pneumonia is almost always accompanied dur- ing the acute stages of the disease by pneumococcus septicemia, and in typhoid fever we now know that the organisms circulate freely in the blood during the first two weeks of the disease, and often longer than this. In these and other conditions the bacteria may be gradually de- stroyed and disappear from the blood stream as the immunity of the subject increases. In other cases the bacterial activities may be partially checked, the process becoming slower and more chronic. This is especially often the case when micro-organisms after entrance to the circulation have found a secondary lodgment upon a heart valve, from which a continuously renewed supply of bacteria can be given off to the blood. A special form of such "malignant endo- carditis" caused by the Streptococcus viridans is particularly apt to take this chronic course. The presence of bacteria in the blood is not, therefore, as for- merly supposed, an invariably fatal condition. Adami's recent work would indicate, moreover, that bacteria may normally enter the portal or even the general circulation from the intestine during health. This condition of "sub-infection," as he calls it, is more fully discussed on p. 234. That colon and other in- 44 Fraenkel. V. Leyden Festschr., 1902. THE PROBLEM OF VIRULENCE 25 testinal bacteria may often penetrate into the portal circulation is indicated by the occasional occurrence of colon bacillus abscesses after trauma of the liver. In most septicemias, however, caused by virulent bacteria the invasion of the blood stream persists, rapid multiplication occurs and leads to death. From the circulation the bacteria may gain lodgment in various organs and cause the formation of secondary abscesses. This condi- tion is known as "pyemia," and may be caused by almost any bac- teria which are capable of producing septicemia. Thus staphylo- cocci, streptococci, or pneumococci may lodge in bones, joints, brain, or kidneys, in fact in any organ in which they can gain a foothold. However, there are evidences of distinct tissue predilections on the part of certain germs. Thus the virus of rabies and that of polio- myelitis, though to some extent universally distributed, seem espe- cially to concentrate in the nervous system; cholera spirilla and dysentery bacilli appear to find conditions most favorable for de- velopment in the intestinal mucosa; amebic abscesses are most common in the liver; gonococcus infections when generalized find secondary localization with particular frequency on heart valves and joints ; leprosy bacilli have a predilection for the nerve sheaths ; and glanders bacilli injected into the peritoneum of a male guinea pig localize with such regularity in the testicles that the experiment has diagnostic value (Strauss test). Conversely it is only explicable on the assumption of such selective lodgment that tubercle bacilli, even though otherwise universally distributed through the body, will be absent from striped muscle tissue, and rare in the walls of the stom- ach. Such selection — as far as we can account for it at all, seems to depend upon the varying cultural conditions encountered by the germs in different organs. On the other hand, localization may also be dependent upon accidental conditions such as trauma. Infections in which the en- trance of bacteria is coincident with injury — as in the case, for in- stance, of compound fractures — will be able to spread throughout the injured region much more easily than they could enter the healthy tissue. In fact, it is well known that local tissue injury at the point of inoculation favors infection since it furnishes a rich substratum for growth in the form of dead cells or blood clot and interferes with the accomplishment of a normal protective reaction. In cases in which bacteria are circulating in the blood mechanical injury may create a focus of reduced resistance on which the in- vaders can gain a foothold. It is in this way perhaps that, among other things, we can explain tuberculosis of joints or bones which present a history of injury preceding the development of the infec- tion— or the pleurisy and lobular pneumonias which have been known to ensue upon the fracture of a rib. It is also possible that bacteria may be distributed in various 26 INFECTION AND RESISTANCE organs directly from the initial focus by embolism or by the massive invasion of a blood vessel. It is by such breaking into a vein that Weigert explains the generalization of miliary tuberculosis. The inflammatory reaction which usually ensues at the point of entrance of bacteria is merely a result of the local struggle between invader and tissues, and the violence of this reaction is in a large measure an indication of the resistance of the infected subject. When, for instance, a streptococcus of moderate virulence gains lodgment in the skin of a healthy individual the rapid mobilization of leukocytic and other defences may prevent further invasion by the bacteria and lead to a struggle which is clinically evidenced by severe local symptoms. Did the virulence of the streptococci far overbalance the powers of resistance the local struggle might be reduced to a minimum, the infection progressing without any, or with but a slight local, reaction. The fact that pneumococci lodging in the human lung ordinarily cause lobar pneumonia is merely an evidence of a considerable degree of resistance to these germs on the part of the average human being. Pneumococci introduced into the pulmonary alveoli of very susceptible animals (rabbits) may pass directly through into the circulation, causing fatal septicemia with- out leading to a more than mild and temporary reaction in the lungs themselves. If, as in Wadsworth's 45 experiments, the rabbits are partially immunized — that is, their resistance increased before the pulmonary inoculation is carried out — a violent local reaction, anal- ogous to lobar pneumonia, may follow, the severity of the reaction at the portal of entry being manifestly an evidence of more energetic opposition to further penetration of the bacteria. The entrance of bacteria into the deeper tissues, and even the circulation, without any, or with but slight, local evidences of infec- tion at the point of entrance is by no means rare. The innocent appearance of the site of the entrance of the bacteria in generalized streptococcus infection is a common surgical observation, and a strep- coccus-infected wound of the hand or leg in a patient dying of septi- cemia may appear but slightly inflamed and edematous and incom- parably milder in appearance than a staphylococcus boil with which the patient is walking about and suffering hardly any systemic dis- turbance. Between the time of entrance of the bacteria into the body and the first appearance of symptoms of disease there is always a definite interval which is spoken of as "incubation time." This period is made up of two definite divisions — one the time necessary for growth, distribution, and accumulation of the bacteria, the other the time necessary for the action of the toxin or poison which may be secreted. The latter, the incubation time of the toxin, is a subject which is still unclear in many of its phases, and will be discussed 45 Wadsworth. Am. Jour, of the Med. Sc., Vol. 27, 1904. THE PROBLEM OF VIRULENCE 27 in the following chapter (see p. 37). The former, however, is easily comprehended, in fact, is to be expected. For the small number of bacteria which gain entrance to the tissues in spontaneous infection is entirely inadequate in itself to produce symptoms. It is neces- sary that multiplication shall take place until the bacteria have ac- cumulated in number sufficient to cause noticeable physiological disturbance. That the interval necessary for this must vary accord- ing to the number of bacteria originally introduced, the virulence of these, and the specific resistance of the patient goes without saying. Von Pirquet and Schick have suggested also that the incubation time may correspond roughly to the interval during which the sub- ject is becoming "allergic" or hypersusceptible to the bacteria or virus. This will be discussed at greater length in the chapter on anaphylaxis.46 But within the limits of the variations introduced by these fac- tors the incubation time of each infectious disease — if spontaneously acquired — is sufficiently uniform to be characteristic. Thus the pri- mary lesion in syphilis follows the inoculation after an interval of two to three weeks, rabies follows inoculation with street virus after ab9ut four to six weeks, the period being somewhat dependent on the location of the bite ; typhoid fever takes about two weeks to develop ; gonorrhea about five to seven days; small-pox about two weeks; yellow fever three to five days; and scarlet fever and diphtheria about two to six days. In general, it may be stated that within the limits observed for each particular infection the shorter the incuba- tion time the more severe is the infection. Thus if tetanus follows inoculation with the tetanus bacillus within seven days the prognosis is far more grave than when the incubation time has occupied two or three weeks. And if localized and general symptoms follow rap- idly (within twenty-four to forty-eight hours) after a streptococcus infection it is likely that the process is a very severe and virulent one. 46 Von Pirquet u. Scbick. Wien. kl Woch., 16, 1903, pp. 758 and 1244. CHAPTER II BACTERIAL POISONS WHEN bacteria have gained a foothold anywhere within the animal body the local and general disturbances which follow, in all but the mildest and most trifling cases, are such that we cannot account for them solely on the basis of mechanical injury. It may well be that the obstruction of capillaries and lymphatics and the pressure upon parenchyma cells, always incident to inflam- matory reactions, contribute materially to local destruction, and thereby indirectly to systemic effects. However, even in diseases like anthrax, in which the body of the victim after death is found flooded throughout with masses of bacteria, these factors cannot fully explain the clinical manifestations. And such cases, indeed, are extreme examples, since, in the large majority of bacterial diseases, the illness resulting in the patient is severe out of all proportion to the extent of the tissue area invaded. Moreover, all infections, if at all severe, whatever their nature or localization, give rise to fever, and this symptom alone, if care- fully observed from hour to hour, may be sufficiently characteristic to indicate the specific micro-organism which is causing the illness. With this there occur alterations of the blood picture, either a numerical increase of white blood cells (leukocytosis) or a change in the relative proportions of the different kinds of leukocytes — or again an anemia caused by the destruction of red cells. There may also be degenerative changes in parenchyma cells of organs far re- moved from the actual site of bacterial lodgment. All these facts indicate very definitely that, apart from localized tissue destruction or purely mechanical interference with function by capillary ob- struction or pressure, there is at the same time an absorption of poisonous substances emanating from the bacteria. From the earliest days of logical investigation into the nature of infectious disease, as soon, in fact, as cultural methods had been introduced, bacteria were studied with the purpose of throwing light upon this phase of their activity. As a result of such investigations Selmi,1 in 1885, described certain basic toxic substances which he obtained from putrefying human cadavers and for which he sug- gested the designation "ptomain" (from Trrw/Aa = dead body). These 1 Selmi. Cited from Hammarsten, "Textbook of Physiol. Chem.," p. 16. 28 BACTERIAL POISONS 29 poisons were later more extensively studied by Brieger,2 Gautier,3 Griffiths,4 and others, and it was at first surmised that the formation of such substances in the infected animal might be held responsible for the toxemic manifestations which accompany bacterial disease.5 This, as we shall see, is not the case. Ptomains are probably not formed in traceable quantity in the living tissues and are not in any way identical with the specific bacterial poisons which are respon- sible for the toxemia of infectious diseases. Nevertheless, they have some pathogenic significance, since they are invariably products of the proteolysis caused by bacteria and can give rise to illness when ingested with putrefying foodstuffs. It is important, therefore, that we discuss them briefly and consider their fundamental distinction from the true bacterial poisons. Whenever dead organic material, meat, fish, vegetable refuse, etc., is left to itself under suitable conditions of moisture and tem- perature, putrefaction sets in. As a result of bacterial growth the protein is broken up and among the intermediate products of such proteolysis ptomains appear. Chemically 6 7 8 these substances are basic nitrogenous compounds which may or may not contain oxygen. Because of their basic and often highly toxic properties they have been spoken of as "animal alkaloids." Many of them contain only C, H, and !N", and are ammonia substitution products. (See Vaughan and Novy, loc. cit., p. 248.) Thus some of the simpler ones are: Methylamin=(CH3) E"H2 Dimethylamin= ( CH3 ) 2 NH Trimethylamin=(CH3)3 N" Among those somewhat more complex are: Putrescin=NH2— CH2— CH2— CH2— CH2— KE2 and Cadaverin=N"H2— CH2— CH2— CH2 — CH2— CH2— KE2 Samuely classifies the ptomains according to their nitrogen con- tents as follows: 1. Those with one nitrogen atom (CgH^N) (CJI13!N") (C10H15N) 2. Those with two nitrogen atoms such as putrescin (C4H12^N"2) and cadaverin (C5H14N2) and 2Brieger. "Die Ptomaine," Berlin, 1885; Virchow's Archiv., Vols. 112 and 115 ; Berl klin. Woch., 1887, 1888. 3 Gautier. Cited after Pick, Bull, de I'acad. de med., 1886. 4 Griffiths. Compt. Rend, de I'acad. des sc., Vol. 113. 5 For a historical outline of our knowledge of these poisons, as well as for a thorough treatment of their nature, see Vaughan and Novy, "Cellular Toxins." 6 For a discussion of the chemistry of the ptomains see Vaughan and Novy, "Cellular Toxins," Lea Bros., Philadelphia, 1902. 7 Also Samuely in Oppenheimer's "Handbuch der Biochemie," Vol. I, pp. 794 et seq. 8 See also Wells, "Chemical Pathology," Saunders, Phila., 1907. 30 INFECTION AND RESISTANCE 3. Those with three nitrogen atoms such as methyl guanidin (02H7N,). 4. Finally there is an important group which contains oxygen, such as the substance sepsin (C5H14N2O2) obtained by Faust from putrefying yeast cells. They are not in all cases protein cleavage products, since bodies of the cholin group, cholin, neurin, and muscarin, the two last named highly toxic, are lecithin derivatives, and Samuely points out that other lipoid cleavage products, always present in decomposing tis- sues, may well contribute to ptomain production in the presence of a source of nitrogen. It is interesting to note also that the vegetable poison muscarin, isolated by Schmiedeberg from mushrooms, is chemically identical with a toxic base found by Brieger in decom- posing fish. The ptomains are not poisonous in every case. The chemically simpler ones like methylamin, di- and trimethylamin possess little or no toxicity. Others chemically more complex — like cadaverin and putrescin — may be capable merely of causing local necrosis, while sepsin, closely related to cadaverin in chemical constitution, but containing oxygen, is a powerful poison which acts violently upon the intestinal blood vessels, causing capillary dilatation, con- gestion, and diapedesis.9 The presence of oxygen seems indeed to be necessary for the development of strong toxicity (Brieger, Vaughan, and Novy). Again, the lecithin derivative, cholin, is but weakly toxic, while neurin is exceedingly poisonous. In putrefying mix- tures these toxic bodies appear on or about the fifth or seventh day after putrefaction sets in, and disappear, by further cleavage, more or less rapidly, yielding less complex nitrogenous substances that are non-toxic. With the limited knowledge regarding bacteria and infectious diseases at the disposal of the earlier investigators it was but natural that the discovery of ptomains in cultures of putrefactive bacteria aroused the suspicion that these bodies were responsible for the toxemia of infectious disease. The search for poisonous substances in pure cultures of patho- genic bacteria was, therefore, assiduously taken up by Brieger and his pupils, and, in truth, ptomains were actually found as products of some of the disease-producing micro-organisms, just as they had been found in the mixed cultures involved in the putrefaction of meat. Thus cadaverin was found in cultures of the cholera spiril- lum, another nitrogenous poison, typhotoxin, in those of typhoid bacilli, and still another in tetanus cultures, all of them producing more or less severe illness when injected into animals. In spite of this evidence, however, we have been forced to con- clude that the ptomains cannot properly be held responsible for bac- 9 Meyer and Gottlieb. "Experim. Pharmakologie," 2d ed., p. 262. BACTERIAL POISONS 31 terial toxemia as manifested in disease. In the first place it is doubtful whether ptomains, in noticeable quantity, are ever produced within the living infected body. Then, again, potent ptomains are produced in culture by many bacteria having absolutely no patho- genic power, while highly pathogenic bacteria may produce little or no ptomains. Ptomain production, moreover, is not specific, since the same ptomains may be produced by many different bacteria or mixtures of bacteria, provided the conditions of nutrient materials and temperature are favorable for growth. We cannot therefore account for bacterial toxemia, in which the poison produced by an individual species is characteristic and invariably the same, under varying cultural and environmental conditions, by the production of ptomains. And even when ptomains are produced in culture fluids by pathogenic bacteria their physiological action is usually quite different from that of the poisons produced by the same micro-organ- isms in the infected subject. Briefly summarized, therefore, the ptomains are poisons elab- orated by all bacteria that are capable of producing protein cleavage, if planted on suitable nutrient materials under conditions favoring growth. The matrix of these poisons is the protein nutriment ; they are not products of intracellular metabolism specifically characteris- tic of the bacteria which produce them. Their importance in the production of disease, therefore, is really an indirect one. They may cause disease if putrid meat or other material is ingested, and with it preformed ptomains, which may be taken in and further elaborated by continued putrefaction in the intestines. This form of meat poisoning, without bacteriological in- vestigation, may be difficult to distinguish from such bacterial forms of meat poisoning as those caused by the Gartner bacillus or the bacillus botulinus. Novy 10 believes that true ptomain poisoning of this kind is rather less frequent than formerly supposed. However, in such cases as those of \7aughan, who isolated a poisonous ptomain "tyrotoxicon" from cheese and milk, their importance seems rea- sonably certain. It is also probable that certain forms of auto- intoxication may be caused by the production in the intestinal ca- nal of ptomains resulting from bacterial putrefaction incident to faulty digestive conditions. It is the antagonism to such intes- tinal putrefaction by the acid production of the bacillus Bul- garicus which is probably the basic cause of any favorable thera- peutic effects which have attended the soured milk therapy of Metchnikoff. Again the growth of saprophytes in necrotic tissues such as gangrenous extremities in diabetes or amputation stumps, may lead to the formation of ptomains which, after absorption, can cause disease. In all such cases the process is one determined by the bacterial putrefaction of dead organic materials, and the absorbed 10 Novy in Osier's "Modern Medicine," Vol. 1, p. 223. 32 INFECTION AND RESISTANCE poisons are not true bacterial toxins, since they do not emanate specifically from the cell substance of the micro-organisms but rather represent incidental cleavage products of the nutrient materials. Therefore, also, the ptomains are unspecific — their formation a com- mon attribute of a large variety of saprophytic organisms, their production, as to quantity and kind, primarily dependent upon the nature of the nutrient materials on which the bacteria are grown. In contradistinction to the ptomains, the specific bacterial poi- sons, in the technical meaning of the term, are substances which are characteristic for each individual species of bacteria and truly the products of bacterial metabolism in that they emanate from the cell itself, either as a secretion or excretion during cell life, or as an inherent element of the cytoplasm liberated after death (or possibly as a cleavage product of the disintegrating bacterial protein).11 They are dependent upon the nature of the culture medium only in so far as this favors or retards the normal development of the micro- organisms. While, therefore, a diphtheria bacillus undoubtedly pro- duces the largest quantities of its specific poison on bouillon suitably prepared for this particular purpose, it will also, in smaller amount, produce qualitatively the same poison on all media on which its growth is free and uninhibited, even on a medium such as that of Uschinsky, which is entirely devoid of proteins. The toxins are, therefore, elements of intracellular metabolism, permanently or transiently constituent parts of the cell body. A specific bacterial toxin was first obtained from the diphtheria bacillus by Roux and Yersin12 in 1889. They discovered that if diphtheria bacilli were grown on veal broth and the cultures filtered through porcelain candles, after seven days at 37.5° C. the filtrates were highly toxic, producing the same symptoms and autopsy find- ings in rabbits, guinea pigs and birds wThich followed the injection of the living bacilli themselves. The poison was therefore a soluble product of the bacteria during the period of their vigorous growth, apparently given up by them to the culture fluid. Very soon after this, in 1891, Kitasato 13 discovered a similar specific toxin in cul- ture filtrates of the tetanus bacillus, and it was the hope of bacteri- ologists that analogous poisons could be determined for all patho- genic bacteria. This hope, however, has been disappointed. It was soon found that cultures of cholera spirilla, typhoid bacilli, and many other germs did not yield toxic filtrates of this kind but that the poisons in these cases seemed to be firmly bound to the bacterial bodies dur- 11 In connection with this read the discussion on anaphylaxis in chapter XVII, p. 413. 12 Roux and Yersin. Ann. de VInst. Pasteur, Vol. 2, 1889. 13 Kitasato. Zeitschr. f. Hyg., 1891, Vol. 10. BACTERIAL POISONS 33 ing life, and given up to the surrounding media only after death and disintegration of the cells. Pfeiffer 14 was the first one to formulate this conception in his studies upon cholera poisons. He found that when cholera spirilla were grown upon broth and filtered after 6 or 7 days, the filtrate was but slightly toxic, but that, in this case, unlike the conditions pre- vailing in diphtheria and tetanus cultures, the residue of bacterial cell bodies, even after they had been killed by chloroform, thymol, or drying, were powerfully poisonous. We have then two main classes of specific bacterial poisons. One — typified by diphtheria and tetanus poisons — is produced during the period of energetic growth by the living bacteria, is given off to the surrounding culture fluid as a secretion or excretion, and can be obtained in bacteria-free filtrates at a time when few, if any, of the micro-organisms have died or disintegrated. These are spoken of as "true toxins" or "exotoxins." The other group — typified by the cholera poisons as described by Pfeiffer — is apparently an intracellular, constituent part of the bac- terial body — not given off during life and not, therefore, obtained in filtrates of young living cultures. If the cultures are preserved until cell death has taken place and the dead bodies have been extracted by the culture fluid, the filtrate becomes gradually more toxic. The bodies of such bacteria are in themselves powerfully toxic when injected, dead or alive. These poisons for obvious reasons Pfeiffer has named the "endotoxins" since he regarded them as specific and definite substances, present as such in the living bacterial cell. In addition to the endotoxins the bacterial protein contains sub- stances which attract and lead to the accumulation of leukocytes. In other words, they exert a positive chemotactic influence. This was first observed in 1884 by Leber,15 who induced the formation of pus by injecting dead staphylococcus cultures, and, later, found that the same effect resulted from the injection of alcoholic extracts of staphylococci. These chemotaxis-inducing substances were later particularly studied by Buchner. Buchner 1G extracted them from many varieties of bacteria, independent of pathogenicity. Although there are quantitative differences, all bacteria seem to contain such substances, and Buchner believed the chemotactic property to be a general attribute of the bacterial protoplasm. He speaks of his ex- tracts as bacterial proteins. The true toxins or exotoxins, then, appear to be products of liv- ing bacteria given off from these very much as are the ferments and enzymes by which micro-organisms cause cleavage of carbohydrates or proteins — and indeed the French school, from the first, compared 14 Pfeiffer. Zeitschr f. Hyg., Vol. II, 1892. 15 Leber. "tiber die Entziindung," Leipzig, 1884. 16 Buchner. Berl. klin. Woch., 1890. 34 INFECTION AND RESISTANCE these toxins to enzymes, with whicn, as we shall see, they have much in common. The endotoxins — on the other hand — at least as con- ceived by Pfeiffer, are structural ingredients of the bacterial proto- plasm which are toxic when brought into solution as the cells break up. Concerning the accuracy of this conception, however, much doubt has recently arisen, as a result of researches which will be discussed below. These two types of poison, moreover, differ from each other not only in mode of origin but in biological characteristics far more fundamental than this. The discovery of diphtheria toxin by Roux and Yersin was fol- lowed by diligent investigations into the toxic properties of all known pathogenic bacteria, and it was soon found that a few only of these germs could produce poisons biologically similar to that found in diphtheria cultures. It was in the course of investigations of this kind, indeed, that Pfeiffer, failing to discover an exotoxin in cultures of cholera and other germs, formulated his endotoxin theory. The list of true toxin or exotoxin producers, then, is short. Among the more important are, in addition to the diphtheria and tetanus bacilli — which have been mentioned above — the Bacillus botulinus,17 the Bacillus pyocyaneus^ and that of symptomatic an- thrax.19 It has also been claimed that similar toxins are formed by the cholera spirillum (Brau and Denier),20 by the dysentery bacillus of the Shiga-Kruse type (Kraus and Doerr) 21 and the Bacillus ty- pliosus (Arima).22 In the cases of the three last-named organisms, however, the secretion of a true exotoxin has not been accepted as a fact by all observers. Indeed, even though such substances may pos- sibly be produced by these bacteria in small amounts it is not likely, in the light of our present knowledge, that they play more than a sec- ondary role in the toxemic manifestations of cholera, dysentery, and typhoid, the important poisons in these cases being those derived from the bacterial cell bodies. Similar in essential properties to the true exotoxins also are the erythrocyte poisons (hemotoxins) produced by many bacteria which cause hemolysis of red cells, and the leukocyte-destroying poison (leukocydin) which is a product of the Staphylococcus aureus. All of these "true bacterial toxins" or exotoxins, apart from sim- ilarity of origin, as soluble secretions of the living bacteria, possess certain common biological characteristics which sharply differentiate 17 Kempner. Zeitschr. f. Hyg., Vol. 26, 1897. 18 Wassermann. Zeitschr. f. Hyg., Vol. 22, 1896. 19 Grassberger and Schattenfroh. Wien Deuticke, 1904. 20 Brau and Denier. Ann. de I'Inst. Past., Vol. 20, 1906. 21 Kraus and Doerr. Wien kl Woch., 42, 1905. 22Arima. Centralbl. f. Bakt., I, Vol. 63, 1912. BACTERIAL POISONS 35 them from the "endotoxins." These characteristics they share with a number of non-bacterial substances such as the vegetable poisons ricin, crotin, and abrin, with animal poisons like snake venom and spider poison (arachnolysin), and, in certain important respects, with the substances spoken of as enzymes. Thus the bacterial true toxins are not biologically unique sub- stances. Both in themselves and in regard to the reactions they elicit when injected into the animal body, they share certain cardinal properties with analogous substances derived from the higher plants and from animals. And it is important to recognize at once that we are dealing here, as in other phases of the study of bacterial immu- nity, with broad biological laws, which find application not only in bacteriology, but in general pathology and in the phenomena of pro- tein metabolism in general. It so happens that these phenomena have been studied and are most easily elucidated in connection with bacteria. But their general significance must not be lost sight of. The cardinal characteristic which unites air of these substances into a single well-defined biological group is their property of in- ducing the formation of antitoxins when injected into animals. This property is so important and its thorough comprehension so essential that we may be permitted to digress briefly in order to make it clear. As we shall see, in subsequent chapters, all substances which lead to the formation of specifically reacting antibodies in the treated animal are spoken of as '^Imfigem* 7)r ""antibody-inducing sub- stances." The class of "antigens" is a large one, including all known proteins, and possibly some of the higher proteid split prod- ucts, and protein-lipoid combinations, though the "antigenic" prop- erties of the last two are still in controversy. But among this large group of substances it is only the bacterial true toxins (exo- toxins), obtained in broth filtrates of living cultures, together with the vegetable poisons and other substances we have classified with them above, which induce in the blood of the treated animal a neutralizing antibody — (antitoxin) — which inhibits quantity for quantity the activity of the injected toxin or vegetable or animal poison. This property of eliciting the production of antitoxin in the animal body alone separates these substances sharply from all other antigens, toxic or otherwise, and, in this respect, they differ sharply from the so-called "endotoxins" against which no antitoxins can be produced. As an important secondary characteristic of this group of sub- stances we may regard their chemically indefinable nature. In the case of none of them have we any definite knowledge of chemical constitution except in so far as it has been hitherto impossible to separate them from the protein molecule. The intensive chemical study of the toxins has universally resulted in failure to obtain a protein-free product which has the characteristic toxic properties of 36 INFECTION AND RESISTANCE the original filtrate, or its antitoxin-inducing power. Concerning the methods which have been employed in the study of the chemistry of these substances we will have more to say in another place.23 It is safe to summarize all this work for our present purposes, by stat- ing that, whatever the method employed, until now all of the prep- arations obtained have given one or another of the protein type- reactions, and that none of them can be positively accepted as pro- tein-free. The results here obtained have been entirely analogous to those obtained in similar investigations upon enzymes. (See also discussion of antigens, chapter 4.) The analogy with enzymes is indeed a striking one and noted by the first investigators of a true toxin, Roux and Yersin. Biolog- ically, of course, we have the cardinal similarity in that the injec- tion of toxins into animals induces the production of antitoxin, and treatment with enzymes induces specific and neutralizing anti-en- zymes. In addition to this, they are alike in their susceptibility to heat (both being destroyed when in solution by temperatures over 80° C.), in their gradual deterioration on standing, and their mys- terious activity in small quantities upon disproportionately larger masses of the substances they attack. There is, however, one impor- tant difference between the two in their mode of action. For, while the toxins are apparently bound or neutralized by the tissues they attack, the action of an enzyme seems rather to be a process in which the enzyme unites with the substance it acts upon, is released as the result is attained, and freed for further action, without noticeable loss of quantity. Such catalytic properties have not yet been satis- factorily demonstrated for the bacterial toxins. However, there are other modifying factors which may account for lack of similarity in this respect, and in all other important points the two classes of sub- stances are closely analogous. The property of heat sensitiveness, which is a characteristic of bacterial exotoxins and enzymes, is shared with them by all of the substances mentioned above except snake venoms. Snake venoms are not destroyed completely until the temperature is raised to 75°- 80° C. The earlier contention of Leclainche and Vallee, that the toxin of symptomatic anthrax possessed similar heat stability has been satisfactorily refuted by Grassberger and Schattenfroh,24 who find that heating it to 50° C. for an hour completely destroys it. There is another important attribute of the true toxin which deserves discussion, though we are by no means in a position to offer any satisfactory explanation for it. We refer to the incubation time which elapses between the administration of a toxin and the occur- 23 An extensive and authoritative summary of this phase of the subject is that of E. Pick in "Kolle u. Wassermann Handbuch," etc., 2d ed., Vol. 1. 24 Grassberger and Schattenfroh. "Tiber das Rauschbrandgift, etc.,1'' Wien. Deuticke, 1904. BACTERIAL POISONS 37 rence of symptoms. Here again snake poisons form an exception — since local manifestations may appear within an extremely short period after the injection of the venom or as the result of a snake bite. However, in the case of all other toxins there is a definite lapse of time between the entrance of th*e poison and the first symp- toms, local or general. This interval is longer when small doses are given — shorter when the doses are large — but is never entirely elim- inated— even when many times the fatal dose is given. In the case of tetanus poison, for instance, injections into a horse may not cause symptoms for as long as four or five days. In mice, animals that are extremely susceptible, the incubation time may be shortened from 36 to 12 hours if we inject 3,600 lethal doses, but, in any case, whatever the dose, this interval cannot be shortened below 8 or 9 hours.25 Many attempts have been made to explain this. Ehrlich, as we shall see, assumes that the action of a poison depends upon two occurrences : one, the union of the poison with the vulnerable cell, the other the gradual injury of the cell by the toxic atom groups in the poison molecule. The time necessary for the institution of this process, he believes, explains the interval. Richet has suggested that the toxin itself may not be potent until acted upon by the body of the recipient and transformed into a potent form. His views are more directly related to the phenom- enon of anaphylaxis and are discussed in another section. De Waele has recently advanced a theory which implies that the incubation time represents the period necessary for the gradual concentration of the poisons in the vulnerable tissues, a process which depends either upon chemical affinities or solubility of the toxins in the cell lipoids. A little at a time would then be absorbed by the vulnerable cells as they come in contact with the poison, through the circulation, and the symptoms would not appear until a definite intracellular concentration had been attained. His views are so closely bound up with the theories on the selective action of the toxins upon individual tissues and organs that they will be rendered clear as we proceed with a discussion of the latter. The majority of pathogenic bacteria do not, as we have seen, pro- duce true toxins or exotoxins. Cultures of cholera spirilla, plague bacilli, and of many other bacteria do not yield toxic filtrates until the cultures have been allowed to stand for prolonged periods during which extraction and possibly autolysis have occurred. In these cases, moreover, definite toxic properties can be demonstrated in the dead cell bodies or in extracts prepared by various methods. In no case, however, is the injection of these "endotoxins" followed by the production of antitoxins. It was very natural to suppose that in micro-organisms of this class the toxic principle might be present in the form of a preformed intracellular poison which could be ex- 25 De Waele. Zeitschr. f. Imm., Vol. 4, 1910. 38 INFECTION AND RESISTANCE tracted or which became free as cell-death occurred and disintegra- tion ensued. It was assumed that, when bacteria entered the animal body and were destroyed by the action of the serum or cells, these endotoxins were liberated and poisoning resulted. The»very protective action of the serum, which prevented the extension of the infectious in- vasion, by limiting bacterial growth, was thus looked upon as the agency by which the endotoxins were set free. Experiments by Radziewsky and others, in which it was shown that large doses of bacteria injected into immunized animals were violently toxic and more rapidly fatal than corresponding amounts injected into normal animals, were taken to mean that in the immune animals a more powerfully cell-destroying property of the serum led to a more rapid liberation of the endotoxins. This was the conception of Pfeiffer and, in more recent theoret- ical discussions, that of Wolff-Eisner. Its essential features con- sisted in the assumption that the poisons were preformed and were contained within the cell body as such, and that they were specific for each micro-organism, determining to a certain extent its pathogenic properties. Thus typhoid endotoxin, cholera endotoxin, or dysen- tery endotoxin was supposed each to possess its own particular pharmacological properties by which the clinical manifestations of the respective diseases were partially determined. It is chiefly the work of Vaughan26 which has begun to throw doubt upon Pfeiffer's original views, in that Vaughan has shown that all proteins, bacterial or otherwise, would yield, upon cleavage with alkalinized alcohol, toxic split products which possessed many of the pharmacological properties of the so-called endotoxins. In fact, Vaughan succeeded in producing, in animals, fever and other symptoms which are generally associated with infection, merely by injecting into them graded quantities of his toxic split products. Following Vaughan, Friedberger succeeded in showing that toxic substances similar to Vaughan's split products are formed when bacteria of various species are subjected to the action of nor- mal or immune sera, and that such poisons were pharmacologically alike and produced with equal ease from pathogenic and non-patho- genic micro-organisms. These phenomena are discussed in greater detail in our section on bacterial anaphylaxis. It is necessary, how- ever, to point out in this place the uncertainty in which these re- searches have left the conception of endotoxins. They suggest that the toxic effects following upon the introduction of pathogenic bac- teria into the animal body are not due to endotoxins, but are rather the result of the action of toxic cleavage products formed in the re- action between blood plasma and bacterial cell. These split products 26 For a complete discussion of Vaughan's work see Vaughan, "Protein Split Products," Lea & Febiger, Phila. and N. Y., 1913. BACTERIAL POISONS 39 are not conceived as specific for individual bacteria but may be formed from all bacterial proteins, both the pathogenic and the non- pathogenic. The differences in pathogenicity between bacteria of this class would then depend entirely upon their powers to invade — not at all upon their possession of individually peculiar cell poisons. The differences in clinical course and toxemic manifestations would be taken to depend entirely upon the accumulation and the distribu- tion of the invading germs, and the consequently variable energy in the production of the toxic split products from them. Considerable experimental evidence has accumulated in favor of this point of view. We will reserve a consideration of this for a later chapter. In order to do injury to the infected individual the bacterial poisons must be produced in such locations that they can easily enter the physiological interior of the body. None of the poisons that have been so far investigated can produce injury when introduced into the alimentary canal. In this location they are, as a rule, de- stroyed, or they pass through without doing harm. Neither diph- theria toxin nor tetanus toxin will produce symptoms when intro- duced intraintestinally.27 28 29 30 Even cholera poison does not pass through the uninjured intestinal wall. Kruse 31 assumes, and Kolle and Schiirmann 32 seem to agree with him, that the absorption of cholera poison does not occur until the intestinal wall has been injured by the actual growth of the living bacteria. Kruse calls attention to experiments by Burgers in which enormous quantities of cholera poison, i. e., 200 cultures of dead or living cholera bacilli, could be administered to healthy guinea pigs and rabbits by mouth without harm in spite of the fact that these animals are definitely susceptible to the poisons and although the poisons are not injured by the intestinal ferments. It is likely therefore that the absorption of poison begins only after the bacteria have extensively invaded the intestinal mucosa and, by injuring tissue, have opened paths for ab- sorption. In thei case of diphtheria probably a similar condition exists in that the localized injury to the mucous membrane at the point of lodgment of the primary infection prepares a portal of entry. The poison of the Bacillus botulinus alone seems to form an exception to this rule,33 since this substance, though apparently a true bacterial toxin, is absorbed directly from the intestinal canal. With most bacteria this problem does not arise, since the poisons are 27 Meyer and Gottlieb. "Exp. Pharmakol " Urban & Schwartzenberg. Berlin, 1911. 28 Ransom. Deutsche med. Woch., No. 8, 1898. 29 Nencki. Centralbl f. Bakt., Vol. 23, 1898. 30 Carriere. Ann. de I'Inst. Past., Vol. 13, 1899. 1 Kruse. "Allgemeine Mikrobiologie," Vogel, Leipzig, 1910, p. 934. 32 Kolle and Schiirmann in "Kolle u. Wassermann Handbuch" 2d Ed., Vol. 4. 33Madsen in "Kraus u. Levaditi, etc.," Vol. 1. 40 INFECTION AND RESISTANCE elaborated within the tissues, where resorption is a necessary result. Like alkaloids and other organic as well as inorganic drugs, the action of many bacterial poisons is largely selective. Most of these poisons may excite inflammatory reactions if concentrated in any part of the body, but, in addition to this, there is a specific distribu- tion after introduction which indicates that the poison goes into selective relationship with certain tissues and cells. This fact is most clearly illustrated by the bacterial hemotoxins which specifi- cally injure the red blood cells of the infected individual and by such substances as the leukocidin produced by the Staphylococcus aureus, a poison which directly and visibly injures the white blood cells. Here the action is specifically aimed at a well-defined variety of body cell. In considering this problem in connection with infectious dis- ease, it is of great importance to distinguish between selective injury by the poisons transported through the body by the lymph, blood, and other channels, on the one hand, and the selective lodgment of the micro-organisms themselves on the other. The latter may occasion- ally depend on local cultural advantages for the particular bacteria in one organ or another, but may just as often be determined by the peculiar manner of entrance to the body which is most suitable for lodgment of the germs in question, and the degree of local resistance at the point of entrance, which determines whether or not the infec- tion shall be locally limited or permitted to invade beyond this point. In the case of a disease like acute anterior poliomyelitis, where our knowledge of the micro-organisms which cause the disease is yet in its infancy, it is impossible to decide whether the injuries noted in the motor areas of the cord and medulla are due to toxins or the lodgment of the germs themselves. In the case of rabies it seems reasonably sure that the micro-organisms themselves select the nervous system. In such instances as the injury of the motor areas by tetanus poison, that of certain peripheral nerves by diphtheria toxin, or even the characteristic lesions of post-syphilitic maladies like tabes, we can be reasonably sure that we are dealing with the specific action of the poisons, independent of actual localized growth of the infectious agents. Diphtheria toxin, after distribution through the body, may act upon many different tissues, as is evident by degenerations in the heart muscle, liver, and kidney, and the petechial hemorrhages in serous surfaces. In addition to this general action, however, there is a very marked selection of certain nerve centers. By Meyer and Gottlieb 34 diphtheria toxin is classed as a specific vascular poison. Its action results in a rapid sinking of the blood pressure with final 34 Meyer and Gottlieb. "Pharmacology Trans. Halsey," Lippincott, 1914, p. 556. BACTERIAL POISONS 41 cardiac death in spite of artificial respiration. These manifestations seem to have a central origin, with particular action upon the vagi and the phrenic nerves. Apparently also the localization of the diphtheritic lesion may influence the selection of individual nerves, the most concentrated action taking place upon the nerves whose endings are distributed in this particular region, for, as Meyer and Ransom35 have shown, this poison, like tetanus toxin, may be ab- sorbed into the nerves directly through the nerve endings. An in- teresting selective action also of diphtheria poison is the apparently specific alteration of the suprarenal glands which is regularly no- ticed, as enlargement and congestion, in diphtheria-infected guinea pigs, and which has been associated by many workers with the char- acteristic drop in blood pressure which accompanies all severe cases of the disease. Abramow 36 has studied this lesion particularly, and believes that it consists in a degeneration and final disappearance of the chromaffin substance and of the medullary cells. He believes that this, together with degeneration of the heart muscle itself, is of great importance in causing the characteristic vascular failure. In botulinus poisoning there is, as Marinesco B7 and Kempner and Pollack 38 have shown, a direct effect upon the cells of the an- terior horns with degenerative changes in the Nissl granules. Tetanus poison, which has been studied extensively by pharma- cologists, shows a very marked affinity for the nervous system, as, in fact, the symptoms of tetanus indicate. Indeed, while many of the bacterial poisons are distributed by the blood stream to the point of final attack, in tetanus the absorption of the toxin from the lesion or the point of injection takes place entirely by the path of the nerves. That this method of poison distribution might be, among others, an important one was suggested as early as 1892 by Bruschettini,39 who found tetanus toxin in the nerves but not in the adjacent muscle and other tissues surrounding the point of subcutaneous injection. Similar results were obtained subsequently by Hans Meyer, whose experiments were confirmed and extended by Marie and Morax.40 Finally Meyer and Ransom 41 furnished complete proof that the pojson was absorbed from the blood and tissues by the peripheral nerve endings alone and was transported centripetally only by the paths of the neurons. The experimental facts elicited may be sum- marized as follows : 35 Meyer and Ransom. Arch, de pharmacodyn., Vol. 15, 1905, also Meyer, Berl klin. Woch., 25 and 26, 1909, also Arch. f. exp. Path. u. Ther., Vol. 60, 1909. 36 Abramow. Zeitschr. f. 1mm., Vol. 15, 1912. 37 Marinesco. Compt. rend, de la soc. de biol., Vol. 3, 1896. 38 Kempner and Pollack. Deutsche med. Woch., 32, 1897. 39 Bruschettini. Riforma medica, 1892. 40 Marie and Morax. Ann. de I'Inst. Past., 1902. 41 Meyer and Ransom. Archiv f. exp. Path. u. Pharm., 49, 1903. 42 INFECTION AND RESISTANCE 1. When tetanus toxin is injected into the thigh muscles of a guinea pig the poison is found at first only in the sciatic nerve of the same side and in the blood. (The determination of poison was made by injecting macerations of the respective tissues into mice.) If examination was delayed until the symptoms had become general- ized, the poison was found in the opposite sciatic, but the muscle bundles, fat, etc., from the vicinity of the injection area were poison- free.42 2. When a nerve is cut poison absorption ceases as soon as axis cylinder degeneration has set in. 3. If the nerve is cut before the poison is injected the distal end contains poison, the proximal end does not. This again shows that the nerve absorbs the toxin not from its capillaries but solely through the end organs. 4. If a nerve which already contains poison is severed, toxin will disappear rapidly from the proximal end, since it no longer obtains a renewed supply from the periphery. 5. If antitoxin is injected into the nerve, above the point of injection, it will successfully bar the way for the ascending toxin. 6. Severing of the spinal cord prevents the passage of the poison from below upward. These facts ascertained in the case of tetanus find their parallel in the phenomena of the distribution of rabic virus 43 as well as in that of poliomyelitis, in both of which there seems to be a progressive centripetal transportation through the nerves. However, in these conditions we are probably dealing not with a poison but with a living virus and, though analogous, the conditions are not entirely comparable. From the practical point of view these facts regarding tetanus may explain the frequent failure of therapeutic success attending the injection of tetanus antitoxin after the symptoms of the disease have set in, since in such cases the poison is already distributed to the nerves and is largely inaccessible to the antitoxin. They also have pointed a way toward a more hopeful therapy, namely, the method of injecting the antiserum directly into the nerves about the point of injury. It is not surprising, however, in view of the stated facts, that even this is unsuccessful when done at too late a time, after a considerable amount of poison has already passed above the point of injection to the spinal centers. Such selective action on the part of the bacterial poisons is en- tirely analogous to the similar specific action of alkaloids, narcotics, 42 In view of our discussion of the importance of fats in the absorption of tetanus toxin, it seems inconsistent that the toxin does not concentrate in fatty as well as in nervous tissues. This Meyer explains by the inactive and poorly vascularized condition of the fat tissues. 43 Di Vestea and Zagari. Fortschr. d. Med., Vol. 6, 1888. BACTERIAL POISONS 43 and other drugs. In order that the poison may act upon a cell we must, of course, assume that it has either chemical or physical affin- ity for this cell. The problem, as many writers have pointed out, is strongly analogous to that of tissue staining. A dye must be able to form a chemical union with the cell or it must be soluble in the cell substance in order to stain it. The chemical difference between cells is a delicate one and not often definable by our present methods. We can obtain an insight into the principles probably underlying selective action only by inference from the relation between the chem- ical constitution of drugs or their physical properties, solubility, etc., and their respective tissue affinities. These problems are difficult and, to a large extent, obscure. They cannot be directly investigated upon bacterial poisons since these are themselves of chemically un- known nature. But the study of drugs of known constitution has revealed certain definite relations of this kind which have furnished analogies from which the general principles of selection in bacterial poisons can be surmised. It is a well-known fact to pharmacologists that there is a definite relation between chemical structure and toxicity. Fraenkel 44 ex- presses it as follows : "By the addition of identical atom groups in an identical manner, similarly acting substances are obtained." He cites the well-known example of curare; whichever the path by which this poison is injected it leaves intact the tissues with which it comes in contact, but after general distribution acts specifically upon the nerve endings. It had been discovered by Brown and Eraser45 that by introducing methyl radicles (CH3-) into molecules of various alkaloids, strychnin, morphin, atropin, and others, sub- stances were obtained which paralyzed nerve endings, and this irre- spective of their previous physiological action. It appears that the combination of four methyl radicles attached to the nitrogen atom (quaternary bases) universally possesses this paralyzing action. Tertiary bases on the other hand lack this property. CH L3 t< Ammonium lose" "Tertiary lose Y *i™< e active in the given case. Thus, altln iigUrsiicli a polyceptor, :rse, is capable of uniting with the complement which activates the lominant complement, it is capable also of union with a number of other comple- meits which have slight or no functional acf.on whatever — the non-dominant com- plements. This opinion is rendered dia- grtmmatic by Ehrlich and Marshall 53, in th* following way : If one carefully considers the reaso.ns advanced for the assumption of the < teice of such polyceptors it does not soem they are sufficiently forcible to/ lead one to desert the much simpler exjj/lanation of Eordeu Related to the problems discussed in ection with t h " production of aantlr»rabo- (c) Dominant Complement. ,, (d) Secondary Complements. ^^Tkf /-VTK3- ' OT 9.n- X -, AIM /-T j; "< Complementophile Groups ot the Amboceptor: (1) for the Dominant Com- plement. (2) for the Secondary Com- plement. f After Ehrlich and Marshall, - Berl. Min. Wocli., No. £5, 1902.) POLYCEPTOR ACCORDING TO EHRLICH AND MARSHALL. (a) Keceptor of the Cell. (b) Haptophore Group of the Amboceptor. ceptors" t i sensi are those which^ have arisen re- garding the ex- istence of "anti- complement" or "anti - alexins." Ehrlich and Morgenroth Claimed that, by the injection of active horse serum into a goat, they had obtained substances in the goat serum which neutralized hG£se com- plement. They believed that the "'^nti- complements" thus produced neutralized the complement by uniting with its hapto- phore group, thus preventing its combina- tion with the "complementophile group" of the amboceptor. This was their con- clusion because they found that the "anti- complementary" serum exerted no protec- tive influence upon sensitized cells, when these were exposed to the serum and then removed, but that it protected against hemolysis when added to the cells together with the complement. There was apparently no union of the pro- tective substance with the "complementophile" group of the ambo- 53 Ehrlich and Mai-shall. Berl. kl Woch., No. 25, 1902. P]HRLICH AND MORGEN- ROTH 7s CONCEPTION OF THE ACTION OF ANTI- COMPLEMENT. A. Scheme of Hemolysis. B. Action of Anticomple- ment upon Hemolysin. b. = blood cell, c. = com- plement, i. = immune body, a. = anticomple- ment. The complernentoids are not included in the scheme, since in this case they are without influence. 158 INFECTION AND RESISTANCE ceptor, but the protecting substance did act in direct antagonism to the complement itself. From the faclmhat similar anticomplements could be produced when inactivated se^um was injected into animals, they concluded that, on inactivationl there was not a complete destruction of th? complement, but that^ during the process of heating the zymophort group of the complement only was- injured, the* "haptophore group,'r y means of which union fa the tissue* elements would take place, iid 4^rough which, therefore, specific antibody production would be incited, remaining intact. h altered complement they speak o: as "complementoid." Bordet has made similar observations upon the production of anV alexins by the injection into animals both of active and of inact.Ve serum, but in the light of further reseai^.hes, which will be diseased in connection with the problems of alexin-rL&spAbii, cAiifj tnose of Mo- reschi and of Gay, we are forced to the conclusion that the existence of true anticomplements- is by no me Vtfiin, and that the older evi- dence in their favor is found to ] vincing at the present time. In the preceding paragra||(i^we l&xe emphasized the. conceptions of the cytolytic phenomena foraml&ted by Ehrlich and his followers, and although we have brought- out, whenever possible, «the objections of other investigators to^uany of these opinions, we have not yet followed out in a systematic manner the reasoning of any of Ehrlich's opponents. In opposition to the views of his school the leading- position has been taken by Bordet, who, after all, furnished in .his investigation? che fundamental facts which have led to a comprehen- sion of the cytolytic processes.* In explaining Bordet's views we can do no better than to follow out his own exposition as set forth in his article, "A General Resume of Immunity," 54 published with a collec- tion of his papers. He expresses himself, in substance, as follows : That the antigen, in the form of bacteria, blood cells, or cells of any other nature, meets in the body of the treated animal a "recep- tor" complex with which it unites is, of course, plain and agreed to by everyone. That the antibody produced by the tissues in response to such union of antigen with receptor is a direct product of the cells containing the receptors is likely. It is by no means certain, how- ever, or, at any rate, it has never been experimentally demonstrated, that, as Ehrlich maintains, the antibody is identical with the original receptor by which the antigen was fixed or anchored to the tissue cell. It might be assumed with equal justice that the cells of the immunized animal could build up a new substance, not identical with the receptors, in consequence of stimulation by the antigen. It is also by no means certain whether the injected antigen reacts with the body cells themselves or with the normal antibodies which we 54 "Studies in Immunity" oy Bordet and collaborators. Gay, Wiley & Sons, N. Y., 1909. BACTERICIDAL PROPERTIES OF BLOOD SERUM 159 know to exist in many cases. Thus the blood serum of goats may normally often contain hemolysins against rabbit corpuscles. Is it not reasonable to suppose that possibly these may furnish the point of attachment and the source of further antibody production when rabbit cells are injected into goats? In criticism of Ehrlich's as- sumption of the mode of action of heat-stable lytic antibody, Bordet very justly maintains that no proof whatever exists of the "ambo- ceptor" nature of this substance. All that is certain is that the stable substance must unite with the antigen before the alexin or complement can exert its action upon it or be fixed by it. There is 110 entirely valid proof of the existence in this anti- body of a "complemento- l phile" and a "cytophile" i N r? ssifsr group, and no satisfactory instance has been observed in which alexin has united w i t h a heat-stable anti- body which has not previ- 'T:Ri ously been united with an antigen.55 All that has been shown is that the an- SCHEMATIC EEPRESENTATION OF BORDET 's VIEW ,. , ,1 .,1 • , CONCERNING THE INABILITY OF COMPLEMENT Tigen, TOgetner witn T0 UNITE WITH EITHER ANTIGEN OR SEN- specific antibody, forms a SITIZER ALONE AND ITS ABILITY TO BE complex which has an FlXED BY THE COMPLEX FORMED WHEN riv - -, . THE ANTIGEN is SENSITIZED. avidity for alexin, a com- Compare this figure with that representing plex which IS "endowed Ehrlich's conception of the same process, with properties of absorp- tion for complement which neither of its constituents alone possesses." Bordet speaks of the "amboceptors," therefore, as "sensitizers," meaning by this that the antigen, by union with its antibody, is sensi- tized to the action of the alexin. The term "sensitizers" in no way, therefore, implies a preconceived notion, experimentally unproved, of the mode of action or structure of the sensitizer. Since we have graphically explained Ehrlich's opinions, a similar diagrammatic representation may be permitted of Bordet's opinion of the same process of union of antigen and heat-stable antibody with the conse- quent development of alexin-fixing property. In this diagram the ability to absorb or unite with complement becomes evident only after a complex has been formed by the union of the two elements, antigen and antibody. The diagram must not be assumed to mean that the notch into which the complement fits symbolized necessarily an "atom group," but merely expresses the idea of "ability to absorb alexin," not assuming that this ability is 55 Refer also to the discussion of the conglutinins at the end of this chapter. 160 INFECTION AND RESISTANCE either chemical affinity by means of a definite atom group or a mere physical change of molecular equilibrium permitting a specific com- plement absorption.56 It will be seen from the preceding that the controversy between Ehrlich's "amboceptor" conception and the "sensitization" idea of Bordet turns largely upon the existence of a so-called complemen- tophile group of the thermostable antibody. For if it were the case that this antibody possessed an atom group which permitted it to unite with alexin, independent of previous union with antigen, it would go far to support Ehrlich's view. One of the strongest argu- ments brought into the field in favor of such an occurrence by Ehr- lich's followers is the phenomenon of Neisser and Wechsberg, which is usually spoken of as "complement deviation" (Komplement Ablen- kung). In order to make the conditions underlying this phenomenon clear, it will be of advantage to consider for a moment the methods of determining quantitatively the amount of bactericidal antibody (sensitizer amboceptor) in any given immune serum, since it was in working with such titrations that Neisser and Wechsberg made their observations. In carrying out such measurements, it is customary to add in series, to constant amounts of bacteria, varying amounts of inacti- vated antiserum and constant amounts of complement or alexin. These mixtures are set away in the thermostat for 3 to 4 hours, are then mixed with agar and plates are poured. The colonies which develop will give an indication of the number of bacteria killed in each mixture when compared with similar plates poured from tubes in which the same original amounts of bacteria had been mixed with alexin alone. The following table will exemplify such a test: Typhoid bacilli Typhoid antiserum inactive Alexin Result in colonies after 3 hours, incubation Constant quantity ". .1 C. C. .07 c. c. Many thousand Constant quantity 01 c. c. .07 c. c. Many thousand Constant quantity .005 c. c. .07 c. c. 150 colonies Constant quantity .001 c. c. .07 c. c. 200 colonies Constant quantity .0005 c. c. .07 c. c. 800 colonies Control I, constant quantity. . .07 c. c. Many thousand Control II constant quantity Many thousand 56 The diagram on page 159, though possibly not expressing with absolute accuracy the idea of sensitization, was devised because it will remove what seem to the writer frequent misconceptions of Bordet's views. Statements are found in the literature which imply (Ehrlich, "Kraus und Levaditi Hand- buch," Vol. 1, p. 8) that Bordet assumes "dass das Komplement direkt an die Zelle angreift," and deny that there is experimental evidence to support this. It is perfectly true that there is no evidence to show such "direktes BACTERICIDAL PROPERTIES OF BLOOD SERUM 161 BDOODD Wfl In this table it is noticeable that, although there has been con- siderable bactericidal action in the mixtures in which 0.005, 0.001, and 0.0005 c. c. of antiserum were used, the mixtures in which as much as 0.1 and 0.01 c. c. were present, and in which one would nat- urally expect a still greater antibacterial action, the contrary occurred. This surprising and curious phenomenon, showing that an excess of antibody could ac- tually be harmful to the functiona- tion of the bacteri- cidal complex, was explained by Neis- ser and Wechsberg by t h e following reasoning. In tests like the one given above a limited amount of bacteria and a 1 e x i n has been mixed with the enormous amount of anti- body represented in the immune serum. Although bacteria can absorb more of this antibody than is necessary for their solution or destruction, nevertheless the higher concentration given in the table will contain quantities of "amboceptor" so far in excess of the amount that can be absorbed that much of it must remain free in the fluid. Now this amboceptor, possessing a complemen- tophile group, is able to anchor complement or alexin as well as that which has become united with the bacteria. In consequence, there being only a limited amount of complement, some of this is deviated from the amboceptor-antigen complexes by the free amboceptor, and js, in consequence, ineffective so far as bactericidal action is con- cerned. In the higher dilutions of the antiserum, in which no such excess is present, the complement will be concentrated upon the "at- tached" or "anchored" amboceptor, and greater efficiency will result. Graphically Neisser and Wechsberg express their idea in the figure which we reproduce. As to the accuracy of the observations of Neisser and Wechsberg there can be no question, and everyone who has occasion to carry out Angreifen" upon the unaltered cell, but there is evidence that this union takes place after the cell has absorbed the antibody, and no satisfactory evidence to show that the thermostable body is an intermediary, that is, forms a link as conceived in the amboceptor idea. COMPLEMENT DEVIATION AS CONCEIVED BY NEISSER AND WECHSBERG. The complement being united to the unbound ambo- ceptor is thereby deviated from the amboceptor, which has gone into relation with the antigen. (After Neisser and Wechsberg, Miinch. med. Woch., 1901, p. 697.) 162 INFECTION AND RESISTANCE bactericidal tests with any frequency is sure to meet with the phe- nomenon again and again. But their explanation, which involves the assumption of union between free sensitizer or amboceptor, and alexin or complement, without the participation of antigen, cannot be accepted since, search as we may, through the extensive experi- mentation that this problem has inspired, there is no instance on record in which indisputable evidence of such an occurrence has been advanced. On the contrary, there is a mass of satisfactory evidence available which indicates clearly that amboceptor or sensi- tizer alone cannot absorb alexin, and the Neisser-Wechsberg explan- ation seems consequently to be merely an interesting and cleverly conceived but improbable possibility. What, then, is the explanation of the diminution of bactericidal effect in the presence of an excess of sensitizer ? We will see that, in the study of agglutinin and precipitin reactions, phenomena exactly analogous to the Neisser-Wechsberg effect have been noticed, in the case of the agglutinins, the so-called "pro-agglutinoid" zone being a case in point. For these phenomena, as well as for that of Neisser and Wechsberg, explanations have been advanced by the Ehrlich school, similar in principle in that they all depend upon more or less arbitrary assumptions regarding affinity between the reacting bodies. Such explanations, though not outside the realm of possibility, have, however, lost much force since it has been recog- nized that the reactions between serum antibodies and their antigens, in general, take place according to laws far more closely analogous to those governing reactions between colloids than to those governing chemical reactions in which the laws of definite proportions can be applied. And, indeed, the reacting substances in antigen-antibody complexes are, beyond doubt, of the nature of colloids. Now, in many precipitations resulting when two colloids are mixed, an ex- cess of one or the other factor will completely inhibit the occurrence of the precipitation; the reaction taking place only when definite proportions between the reacting bodies are present. The occurrence of such inhibition zones, due to an excessive concentration of one reagent, can be shown for agglutination and precipitation, exactly as it can in ordinary colloidal reactions, and it is more than likely that the Neisser-Wechsberg phenomenon is merely an example of a similar phenomenon. Looked at from this point of view, far from supporting the sup- position of a separate complementophile group and therefore of the "amboceptor" nature of the heat-stable lytic antibody, the Neisser- Wechsberg phenomenon indeed becomes rather a strong argument in favor of Bordet's views, and against those of Ehrlich. For, by introducing the analogy between the lytic and bactericidal processes with colloidal reactions, it takes away much force from the supposi- tion that antigen-sensitizer alexin reactions take place according to BACTERICIDAL PROPERTIES OF BLOOD SERUM 163 laws of definite proportion, an idea which still underlies, though somewhat loosely, many of the more important views of antigen- antibody reactions as conceived on the basis of the amboceptor theory. Gay has suggested also that the Neisser-Wechsberg phenomenon may well be explicable on the basis of the fixation of complement by precipitates. In a succeeding section we will discuss the fixation of alexin, which occurs when a dissolved protein is brought together with its specific antiserum. It is not impossible that this may occur when bacterial emulsions, from which a small amount of bacterial protein may well go into solution, are brought together with anti- serum in concentration. Under such conditions a reaction might readily occur which would lead to the fixation of alexin and its con- sequent deviation from the sensitized bacteria. Of all explanations considered, therefore, that of Neisser and Wechsberg seems to be the least likely. It would seem to us that Bordet's interpretation of these facts is borne out indirectly by cer- tain experiments of Morgenroth and Sachs 5T themselves, in which the mutual quantitative relations between complement and "ambo- ceptor" were studied. In these experiments it was shown that the more highly cells were sensitized, the smaller was the quantity of complement which was needed for their hemolysis, and vice versa, the less the sensitization (the smaller the quantity of amboceptor) the more complement was necessary to produce the same result. The following extract from one of their protocols will illustrate this : BEEF BLOOD CELLS 5%, 1 C. C., ANTIBEEF GOAT SERUM, GUINEA-PIG COMPLEMENT Amount of amboceptor Relative amount of amboceptor Amount of complement for complete hemolysis .05 1 .008 .2 4 .0025 4 8 .0014 A similar relation may be observed by all who have occasion to work with hemolytic reactions. In the present connection this seems to bear out Bordet's interpretation, since, knowing the differences in functional efficiency of various complements for different hemolytic and bactericidal complexes, we could well expect that insufficient sensitization of a red cell or bacterial antigen, not particularly amen- able to the complement employed, might fail to absorb it completely out of the serum, thus giving a negative result which would simulate complete lack of affinity. This research of Morgenroth and Sachs seems further of funda- 57 Morgenroth and Sachs. Berl kl. Woch., No. 35, 1902. 164 INFECTION AND RESISTANCE mental importance in its contradiction of the regularly progressive quantitative relations which strict adherence to the "armSoceptor" idea would seem to impose. The quantitative relations here outlined have been diagrammat- ically represented hy Noguchi as follows : 20 units of A/nboceofor used /n eacfi comti/natton Green . Complement Purple — Amboceptor fad • /feemo/sij /unit of Amboceptor used /n each w/rn various fractions ofa complement I NOGUCHI 's DIAGRAM ILLUSTRATING THE QUANTITATIVE EELATIONS BETWEEN ANTI- GEN, AMBOCEPTOR AND COMPLEMENT. (Taken from Noguchi, "Serum Diagnosis of Syphilis," Lippincott, Philadelphia, 1910.) The essential point of difference between the opinions of Ehrlich and Bordet concerning the processes of hemolysis and bacteriolysis lies, as we have seen, in the conception of the union of alexin or com- plement with amboceptor or sensitizer. Although Ehrlich and his followers admit that the union of complement with amboceptor does not usually occur unless the amboceptor has previously united with the antigen, they still maintain that this may occasionally take place BACTERICIDAL PROPERTIES OF BLOOD SERUM 165 in the case of special complexes in which the complement may di- rectly unite with free amboceptor. This, we have seen, is the basis of the ISTeisser-Wechsberg conception of complement — "Ablenkung" or deviation, and of other ramifications of this theory. Bordet, on the other hand, consistently holds that alexin or complement is at- tached only by the complex antigen-sensitizer (antigen-amboceptor). In the controversy which this difference aroused, an observation was reported by Ehrlich and Sachs,58 which seemed to represent, as they themselves express it, an "Experimentum Crucis" proving Ehrlich' s contention of the intermediary function of the amboceptor in con- trast to Bordet's "sensitization" idea. The facts, as they record them, are as follows: When fresh horse serum is added to guinea pig corpuscles, slight hemolysis results. When inactivated ox serum alone is added to such corpuscles, of course no hemolysis results. If the corpuscles are, on the other hand, exposed to the action of the inactive ox serum, together with fresh horse serum, very active hem- olysis is brought about. Apparently the ox serum sensitizes (or furnishes amboceptor to) the guinea pig corpuscles, rendering them amenable to the action of the complement in the fresh horse serum. In other words, inactivated ox serum can be reactivated by the addi- tion of fresh horse serum. From this one would expect that if the guinea pig cells were exposed to inactive ox serum, then separated from the serum by centrifugalization and fresh horse serum subse- quently added, hemolysis would ensue. However, this was not the case. When the cells were so treated it was found that they had not been sensitized, and, what is more, it could be shown that the ox serum so employed had lost none of its ability to produce strong hemolysis when added to another complex of cells and fresh horse serum. Ehrlich and Sachs concluded that this experiment defi- nitely showed the ability of the amboceptor in the ox serum to unite with alexin independently. The relation to the cell occurred only after the union of the amboceptor in the ox serum and the complement in the horse serum had been established, and if their interpretation is correct, of course, it constitutes strong evidence against the general principle of "sensitization" as conceived by Bordet. This apparent inability of the corpuscles to absorb amboceptor independently out of the inactivated ox serum, and the fact that hemolysis results only if the corpuscles, ox serum, and fresh horse serum are all simultaneously present, are extraordinary and not at all in keeping with the preceding work of Ehrlich and Morgenroth, and indeed with experience of these phenomena in general. It is log- ical therefore to examine more closely the peculiar conditions main- tained in these experiments before applying the reasoning deduced from obviously different phenomena to their explanation. 58 Ehrlich and Sachs. Berl. Win. Woch., No. 21, 1902. 166 INFECTION AND RESISTANCE Bordet and Gay59 accordingly studied the Ehrlich-Sachs phe- nomenon carefully and obtained results which confirmed the experi- mental data of these writers but cast much doubt upon the validity of their conclusions. In going over the experiments of Ehrlich and Sachs, Bordet and Gay made an observation which had apparently escaped the atten- tion of the former investigators. Heated bovine serum has but a slight agglutinating power for guinea pig corpuscles. Fresh horse serum agglutinates them only slightly and slowly. On the other hand a mixture of the two sera agglutinates them very rapidly and completely. The bovine serum apparently possessed an accelerating or fortifying influence both upon the weakly active normal hemoly- sins and agglutinins in the horse serum. Bordet and Gay conse- quently suspected that this property might be due to an undescribed substance, peculiar to the bovine serum. To eliminate the uncertain elements obtaining in experiments in which normal sensitizer is used they now experimented with guinea pig corpuscles, anti-guinea pig sensitizer (from a rabbit immunized with guinea pig blood cells) and guinea pig alexin. They found that sensitized guinea pig cells are hemolyzed by guinea pig alexin very slowly and imperfectly, as is often the case when the alexin comes from the same animal species as the cells. When heated bovine serum was added to the complex of sensitized cells and alexin, rapid agglutination and hemolysis resulted. Their experiments may be tabulated as follows : 1. Cells + guinea pig alexin + heated bovine serum = no agglutination; very slight hemolysis on next day. 2. Cells + sensitizer -+- heated bovine serum = slight agglutination; no hemolysis. 3. Cells + sensitizer + alexin + bovine serum = powerful agglutination and complete hemolysis in 10 minutes. 4. Cells + sensitizer + alexin = very slight agglutination and incomplete hemolysis in 30 minutes. 5. Cells + sensitizer = slight agglutination; no hemolysis. In tube (1) the slight hemolysis was due to the small amount of normal sensitizer present in the bovine serum, and the slight agglu- tination in tube (5) is referable to the agglutinating power of the sensitizer. In tube (3) we see the powerfully accelerating effects exerted both upon agglutination and hemolysis when bovine serum acts upon sensitized corpuscles in the presence of alexin. Bordet and Gay's interpretation of the Ehrlich-Sachs phenom- enon, in the light of these new experiments then, is, in their own words, as follows: "When guinea pig corpuscles are added to a mixture of the two sera they are affected by the sensitizer of the horse serum and, to a certain extent, by the sensitizer in the heated 59 Bordet and Gay. Ann. de I'Inst. Past., Vol. 20, 1906, p. 467. BACTERICIDAL PROPERTIES OF BLOOD SERUM 167 bovine serum. This second sensitizer is, however, superfluous. Its presence is by no means necessary for the experiment. When this sensitization is effected the corpuscles are then in condition to fix the horse alexin. This alexin, however, has only slight hemolytic? power. But once the corpuscles have become sensitized and laden with alexin they are modified in their properties of molecular ad- hesion to such an extent that they become able to attract a colloidal substance of bovine serum, which unites with them. The adhesion of this new substance produces two results : it causes the blood corpuscles to be more easily destroyed by alexin and also agglutinates them energetically. Consequently, a powerful clumping, followed by hemolysis, is observed." Bordet and Gay, therefore, assume that the action of the bovine serum is due to a new substance which they speak of as "bovine colloid." This substance resists heating to 56° C., is probably al- buminous, and has the property of uniting with cells that are laden with sensitizer and alexin, but remains free in the presence of nor- mal or merely sensitized cells. They fortify this opinion by showing experimentally that the "colloid" is removed from bovine serum by absorption with sensi- tized bovine corpuscles which have been treated with horse alexin.60 Bordet and Streng61 later studied this "colloid" more thor- oughly and have suggested for it the name "conglutinin." Streng62 later showed that the agglutinating action of this substance could be shown not only for sensitized and "alexinized" red blood cells, but also for similarly treated bacteria, and that conglutinins were pres- ent not only in bovine serum, but in that of goats, sheep, antelopes, and a number of other herbivores, but apparently absent in cats, dogs, guinea pigs, and birds. The body described by these workers as conglutinin is probably identical with a similar heat-stable serum component reported by Manwaring68 and called by him "auxilysin." 60 Browning (Wien kl. Wochenschr., 1906) had shown that horse alexin may be absorbed by sensitized beef cells without causing hemolysis. 61 Bordet and Streng. Centralbl. f. Bakt., Orig. Vol. 49, 1909. 62 Streng. Zeitschr. /. Immunitatsforsch., Orig. Vol. 2, 1909, p. 415. 63 Manwaring. Centralbl. f. Bakt., 1906 ; Orig. Vol. 42. CHAPTER VII FURTHER DEVELOPMENT OF OUR KNOWLEDGE CON- CERNING COMPLEMENT OR ALEXIN. COMPLE- MENT FIXATION IT will be remembered that Buchner in his first studies upon the "alexin" compared its action to that of an enzyme or ferment, and suggested that the source of this substance might possibly be found in the white blood cells. This thought was very obviously suggested by the observation that bacteria were destroyed within the white blood cells, after phagocytosis, by a process analogous in many ways to that by which they were destroyed by the serum constituents. Hankin,1 in an elaborate study dealing with the problem, maintained the leukocytic origin of alexin on the basis of the observation that increased bactericidal properties closely followed upon the heels of periods of leukocytosis. He assigned the particular property of alexin production to the eosinophile cells, proposing for them the designation "alexocytes." Further study, however, has not justified such an association with the eosinophiles, and Hankin' s opinion has not been experimentally upheld. After Hankin the problem occupied the attention of a number of other investigators, and many of them succeeded in showing that there was, indeed, an increased bactericidal power in exudates rich in leukocytes, and further that bactericidal substances could be directly extracted from leukocytic emulsions. We refer particularly to the early work of Denys and Havet,2 of Hahn,3 of Van de Velde,4 and others, studies which will be described in our chapter on phagocytosis. This work was done before the complex nature of the bactericidal constituents of serum had been demonstrated and before the work of Schattenfroh and others had shown that the bactericidal substances extracted from leukocytes were of a nature quite distinct from the active elements of the serum, and were independent of the participation of alexin. Although these earlier investigations cannot properly be regarded, therefore, as proving the leukocytic origin o£ 1 Hankin. Centralbl f. Bakt., Vol. 12, 1892. 2 Denys and Havet. La Cellule, Vol. 10, 1894. 3 Hahn. Archiv f. Hyg., Vol. 25, 1895. * Van de Velde. La Cellule, Vol. 10, 1894. 168 FURTHER DEVELOPMENT OF KNOWLEDGE 169 alexin, Metchnikoff and his school have nevertheless adhered to this conception for various additional reasons. Metchnikoff distinguishes between two kinds of alexin — the microcytase, which is the bactericidal complement or alexin, and is supposed to originate from the microphages or polynuclear leu- kocytes, and the macrocyiase, which represents the hemolytic and cytolytic alexin or complement, and originates from the mononuclear cells or macrophages. As in the case of the bactericidal alexin, ex- traction methods have been employed to demonstrate that the hemo- lytic alexin took its origin in the macrophages, and at Metchnikoff's suggestion, Tarassewitch 5 prepared hemolytic substances by extract- ing spleen tissue and other "macrophagic organs" in various ways. Here again the identity of the hemolytic extracts with serum hemoly- sins has been placed in doubt. Korschun and Morgenroth 6 have shown that the hemolytic organ extracts were heat stable and alcohol soluble ; Donath and Landsteiner,7 and others, have obtained similar results. It would be quite thankless to review the extensive literature which has accumulated upon this point. It would seem, in summariz- ing it, that no definite proof of the presence of true, active alexin, either hemolytic or bactericidal, within the leukocyte or mononuclear cells has been brought by methods of extraction, and the apparently positive results reported by earlier observers are adequately explained by the discovery of the heat-stable and non-reactivable bactericidal and hemolytic substances in extracts of such cells by Schattenfroh, Korschun and Morgenroth, and many others. It appears, moreover, from these investigations that probably the intracellular substances by which the digestion of ingested bacteria or blood cells is brought about are of a nature entirely distinct from that of the serum anti- bodies and alexins. A very ingenious demonstration of this is found in an experiment first made by Neufeld. Neufeld 8 allowed leu- kocytes to take up highly sensitized red cells. Instead of undergoing prompt hemolysis, as they would if small amounts! of alexin had been added, they were slowly broken up without hemolysis, fragments of hemoglobin remaining after complete morphological disintegration of the erythrocytes. At no time were intraphagocytic "shadow'* forms observed. The failure to extract alexins from dead leukocytes does not, however, preclude the possibility of the secretion of alexins by living leukocytes. This point is one which is, of course, much more diffi- cult to investigate directly. Indirectly the increased bactericidal properties of exudates rich in leukocytes, as found by Denys and Havet, would point in this direction. However, even this is not « 5 Tarassewitch. Cited from Metchnikoff. 6 Korschun and Morgenroth. Berl. kl. Woch., No. 37, 1902. 7 Donath and Landsteiner. Wien. kl Rundschau, Vol. 40, 1902. 8 Neufeld. Arb. a. d. kais. Gesundheitsamt., Vol. 28, 1908, p. 125. 170 INFECTION AND RESISTANCE conclusive, since at the time when these investigations were carried out no discrimination was made between the bactericidal serum sub- stances and those other "endolysins" which might well have been extracted from the accumulated white blood cells. The writer some years ago attempted to approach this problem directly by keeping leukocytes alive in inactivated serum and in Kinger's solution at 37.5° C. for several days in the hope that, after 48 hours, alexin, hemolytic or bactericidal, might appear in these fluids. The experi- ments were entirely negative, but were regarded as inconclusive, since it was impossible to determine accurately how long, or in what pro- portion, the leukocytes had remained alive. One of the basic premises of MetchnikofFs theory on the nature of alexin consists in the conception that alexin is not found in the circulating blood plasma, but appears only when there has been leu- kocytic injury, as in the clotting of blood or in the "phagolysis" which, as we have seen in the chapter on phagocytosis, usually occurs after foreign substances have been injected into the peritoneum, pre- ceding a local accumulation of leukocytes. This point of view seems to be rendered improbable because of the rapid hemolysis which occurs when we inject sensitized red blood cells into the circulation of an animal, but we might here, too, assume a preliminary injury to white blood cells resulting from the intravenous injection. Much less likely to be accompanied by cell .injury is the method of obtaining blood serum by creating an area of artificial edema by ligating a limb — or, as in MetchnikofPs 9 10 experiments, the ear of a rabbit. And, indeed, in edema fluids so obtained little or no alexin is ordinarily found. This fact has been interpreted in favor of MetchnikofFs views, as has also the curious absence of alexin in the aqueous humor of the anterior chamber of the eye.11 12 In this fluid no alexin is present under normal conditions, but if puncture is prac- ticed, and the fluid again taken after a period of three or four hours, alexin is now found, probably, according to Metclmikoffs school, be- cause of the coincident entrance of leukocytes into this space. It is conceivable, however, that the aqueous humor may be free from alexin for other reasons than the absence of leukocytes ; and an injury which is followed by the invasion of leukocytes is pretty sure to be followed also by the entrance of the fluid elements of the blood; i. e., alexin. Much experimental work has been done in which it has been attempted to demonstrate directly that the blood plasma contains no complement or alexin. The most important investigation of this 9 Metchnikoff. Ann. Past., Vol. 9, 1895. 10Bordet. Ann. Past., Vol. 9, 1895. 11 Metchnikoff. Loc. cit. 12 Mesnil. Ann. Past., Vol. 10. FURTHER DEVELOPMENT OF KNOWLEDGE 171 kind is that carried out by Gengou13 in 1001. It was Gengou's primary purpose to obtain the plasma of mammals in such a way that no cell injury would occur. This he accomplished by special methods in which coagulation was avoided without the addition of foreign anticoagulants like hirudin, etc. His technique was, in essence, as follows : He took the blood directly through a paraffined cannula into tubes that had been coated with paraffin, and centrifu- galized it at low temperatures until cell free. This plasma, taken from the paraffin tubes, quickly clotted, and the material with which the experiments were done actually consisted of blood serum. Upon examining the serum so obtained, he found that it exerted practically no bactericidal action. As a result of this investigation he claims to have demonstrated the truth of Metchnikoff's contention that the circulating blood plasma contains no alexin. If borne out, it is true that Gengou's results would very power- fully support this theory, and for this reason a large number of ex- periments have been made since then, with the same end in view. In all such investigations the technical procedures are extremely difficult and, as Addis 14 has recently said, in our opinion quite cor- rectly, it would be impossible to carry out bacteriolytic or hemolytic experiments with mammalian paraffin plasma without obtaining coagulation, and for this reason most of the writers who have re- peated Gengou's experiments have worked, as did he, not with plasma, but with serum. Falloise,15 following Gengou's method exactly, obtained results diametrically opposed to those of Gengou; Schneider,16 also with the same technique, failed to confirm Gengou's results; Herman,17 on the other hand, confirms Gengou. In order to overcome the technical difficulties encountered in working with mammalian plasma a number of writers have more recently experimented with bird blood, which, as is well known, coagulates much more easily than does mammalian blood. Hewlett,18 who worked with goose plasma and peptone plasma, could not con- firm Gengou's results. Lambotte,19 examining the plasma of chick- ens, found no difference between the serum and plasma in their con- tents of bactericidal alexin, as measured against cholera spirilla. Von Dungern, working with fish plasma, obtained similarly negative results, and recently Addis, in a careful comparative study of chicken plasma, found no evidence of differences between plasma and serum in either the bactericidal or the hemolytic alexin. As far 13 Gengou. Ann. de I'Inst. Past., Vol. 15, 1901. 14 Addis. Journ. of Inf. Dis., Vol. 10, 1912. 15 Falloise. Bull, de I'Acad. Eoy. de Med,, 1905, p. 230. 16 Schneider. Archiv f. Hyg., 1908, Vol. 65, p. 305. 17 Herman. Bull, de I'Acad. Eoy. de Med., 1904, p. 157. 18 Hewlett. Archiv f. exp. Path. u. Pharmk., 1903, Vol. 49, p. 307. 19 Lambotte. Centralbl. /. Bakt., I, Orig., 1903, Vol. 34, p. 453. 172 INFECTION AND RESISTANCE as we can tell at present, therefore, we cannot accept, as conclusively proven, the contention that the circulating plasma contains no alexin. Nevertheless the Metchnikoff school have not been discouraged by the various contradictions of Gengou's work, found in the experiments we have enumerated, because they are not satisfied that the technique of (other workers has conclusively excluded cell injury. Owing to the great difficulties of investigations of this kind, when carried out with mammalian blood, it is not impossible that they are justified in this, but nevertheless the assumption of the absence of alexin in the plasma finds so many objections in other observations that the bur- den of proof would certainly rest with Gengou and his supporters. Not the least important of these objections, it seems to us, is based on the very simple experiment of injecting bacteria into the veins of a living animal and finding a very rapid and active phagocytosis. And considering the very probable participation of alexin in the opsonic functions this would seem to point strongly toward the pres- ence of these substances in the circulating blood. The evidence also furnished by the recent developments of our understanding of ana- phylaxis would further tend to strengthen our belief in the presence of alexin or complement in the normal circulation. For, in the process, as we shall see in a later chapter, complement plays an important role. When 3 per cent, salt solution is administered (as in Friedberger's experiments), and the action of complement is thereby inhibited, anaphylactic shock may be greatly diminished. It has also been claimed, chiefly by Walker 20 and by Henderson Smith,21 that, as serum stands upon the clot it at first gains in alexin or complement contents, an occurrence which they attribute to the liberation of alexin from the leukocytes. This observation has not been universally borne out and, even were it unquestionable, it might be dependent upon any one of the numerous factors involved in the complicated process of coagulation rather than upon leukocytic changes only. The failure to obtain definite proof of the origin of alexin from the white blood cells has led to search for the source of these sub- stances in various organs. An interesting series of investigations on this subject are those of Mile. Louise Fassin,22 who believes that she has found reasons for definitely associating the thyroid gland with alexin production. She found that the subcutaneous injection of thyroid extract into dogs and rabbits was followed by a rapid increase of alexin, both hemolytic and bactericidal, and that the same thing was true when thyroid substance was administered by mouth. When the thyroid gland was removed from rabbits a reduc- tion of alexin resulted. Although important, these researches do not 20 Walker. Journ. of Hyg., Vol. 3, 1903. 21 Smith. Proc. Roy. Soc., Series B, Vol. 79, 1906. 22 Louise Fassin. C. R. de Soc. Biol., Vol. 62, 1907. FURTHER DEVELOPMENT OF KNOWLEDGE 173 necessarily prove that the thyroid can be looked upon as a source of alexin, and, indeed, Fassin gives experimental results without drawing any very sweeping conclusions. It might well be that the thyroid secretion is simply concerned in stimulating the production of alexin from another source. Marbe 23 has similarly associated the thyroid gland with the production of opsonins, which, when we consider the probable identity of alexin and normal opsonin, may be taken as a confirmation of Fassin's work. Of great interest also are the series of investigations which asso- ciate the liver with the production of alexin. The basis of such investigations is found in the observations made by Morgenroth and Ehrlich 24 that there is a diminished production of complement or alexin in dogs subjected to phosphorus poisoning, with consequent degeneration of the liver. The first investigator to study this ques- tion experimentally was Xolf.25 Nolf tried to approach it by extir- pating the liver in dogs, and found that his results were unreliable by this method. He then experimented with rabbits and found that when the liver was extirpated in these animals and the vena cava anastomosed with the portal vein (Eck fistula) the animals would survive for three or four hours. This period, though short, was suffi- cient to show definite changes in the blood. Taken just before death it differed from that taken just before the operation in a number of important respects. There was relative incoagulability, there was autohemolysis, and with these there occurred an extreme fall of alexin or complement. Serious objections may be brought against ISTolf's experiments. In the first place the operation as performed by him results in shock and injury so profound that rapid death ensues, conditions under which not only the complement-producing functions but all functions, secretory and otherwise, are reduced. Miiller 26 ob- jects to Golf's experiments chiefly for the reason that he did not pre- vent the absorption of toxic substances from the intestine, materials which could now enter the general circulation without any longer be- ing neutralized by the liver functions. Miiller, for this reason, re- peated Golf's work but, by a complicated technique, temporarily shut off the intestinal circulation in addition to extirpation of the liver. He found, in agreement with Xolf, that exclusion of the liver from the circulation resulted in the prompt diminution of complement or alexin. In all such experiments, however, the very profound shock which necessarily occurs in the animals would seem to us to vitiate the results. Moreover, Liefmann 27 has repeated Miiller' s experi- ments without being able to obtain the same results. Not satisfied 23 Marbe. C. E. de la Soc. Biol., Vols. 64, et seq., 1908-1909. 24 Morgenroth and Ehrlich. In Ehrlich's "Gesammelte Arb.," etc. 25 Nolf. Bull, de I'Acad. de Science de Belg., 1908. 26 Miiller. Centralbl f. Bakt., Vol. 57, 1911. 27 Liefmann. Weichhart's Jahresbericht, Vol. 8, 1912, p. 155. 174 INFECTION AND RESISTANCE with these experiments, however, Liefmann experimented on frogs; in whom, as Friedberger has shown, extirpation of the liver is not so rapidly fatal as in warm-blooded animals. He removed the livers of frogs in a number of cases and, although his animals lived about a week, there was no definite diminution of the hemolytic properties of the serum. It seems, therefore, that the origin of alexin in the body is by no means settled and requires further investigation. Equally unsatisfactory have been the attempts to define the chem- ical nature of the complement or alexin. In the investigations deal- ing with the hemolytic action of cobra venom it seemed at first as though a clue to this problem had been found. Flexner and Nogu- chi 28 made the interesting observation that cobra poison alone does not hemolyze the blood cells of certain animals, namely those of cattle, goats, or sheep, if these cells are washed entirely free of serum. This seemed to suggest that the serum of these animals con- tained some activating substance. It also seemed to indicate that the cells of other animals, which were easily hemolyzed, even when en- tirely freed of serum, might contain such an activating substance within themselves. The behavior of this activating substance toward snake-venom hemolysis was therefore very similar to the action of complement, except in one important respect, namely, as Calmette 29 showed, almost all sera were rendered more efficient for the activa- tion of snake venom when heated to 65° C., whereas complementary properties of sera for other hemolyzing complexes are, of course, destroyed at 56° C. Kyes,30 31 on further studying these phenomena, extracted the red blood cells of rabbits and other animals whose cells were hemolyzed by snake venom alone, by shaking them up with distilled water, and showed that, with these extracts, he could activate the venom against ox, goat, and sheep corpuscles, cells which were not ordinarily hemolyzed by the venom without the addition of serum. Similar activation of the venom with extracts of the ox, goat, or sheep corpuscles was not possible. He concluded from this that the blood cells of the rabbit, dog, guinea pig, and man pos- sessed an "endocomplement" for the snake venom; that is, a com- plementary substance contained within the cells, while in the other species it was found in the activating serum only. The thermostability of such venom "complements" encouraged him to attempt their isolation, and he found that they were ether- soluble, indicating their lipoidal nature; and, finally, after several negative attempts with activation by other lipoids, he determined that lecithin, added to the corpuscles and the snake venom, brought 28 Flexner and Noguchi. Journ. Exp. Med., Vol. 6, 1902 ; Univ. Pa. Med. Bull, 1902 and 1903. 29 Calmette. C. R. de VAcad. des Sciences, p. 134, 1902. 30 Kyes. Berl klin. Woch., Nos. 38 and 39, 1902. 31 Kyes and Sachs. Berl. klin. Woch., Nos. 2-4, 1903. FURTHER DEVELOPMENT OF KNOWLEDGE 175 about a rapid hemolysis. This seemed to explain both why heated serum could activate the venom in some cases, and why some varie- ties of blood cells could be hemolyzed without serum, since lecithin is a substance widely distributed both in the fluids and cells of the animal body. His further studies seemed to show that, by proper chemical manipulation (bringing together cobra poison with lecithin in chloroform solution), he could produce a combination of the two which he called acobra lecithid," a substance which apparently "acti- vated" cobra venom. He conceived it as the "amboceptor-comple- ment" complex of the cobra hemolysin, which acted hemolytically upon all varieties of blood cells. These researches of Kyes aroused much interest, chiefly because they seemed to furnish an example of a chemically definable com- plement, lipoidal in its constitution. Recent researches by Von Dungern and Coca,32 however, seem to prove that, while Kyes' ex- perimental facts were perfectly accurate, his conclusions do not seem to have been warranted. Yon Dungern and Coca showed that the cobra venom contains a lipoid-splitting ferment which acts upon the lecithin, liberating substances from it which hemolyze in the same way as do many other non-specific substances. The cobra-lecithid, according to this, would represent merely a lecithin derivative which happens to have hemolytic action without any specific relationship to the hemolytic properties of the venom itself. Thus, even in this case, unfortunately, we are not in possession of facts which bring 'us nearer to a chemical understanding of the complementary substances or alexins. In the further development of attempts to define alexin or com- plement chemically, two further researches are of importance, namely, those of Von Liebermann 33 and of Noguchi. In both in- vestigations it is suggested that the alexin may consist of a combina- tion of soaps and proteins. Noguchi 34 showed that the hemolytic organ-extracts described by various observers were soaps, a possi- bility which had been previously considered by Sachs and Kyes.35 Noguchi further established analogies between his soaps and com- plement as follows: Sensitized blood cells are hemolyzed by mix- tures of soaps and inactivated guinea pig serum, while normal ery- throcytes are not hemolyzed by similar mixtures; furthermore, like normal complement, such serum-soap mixtures are inactivated by prolonged preservation and by heating at 56° C. Objections were soon made to the findings of both Xoguchi and Von Liebermann by Hecker,36 whose experiments seemed to show that when sensitized 82 Von Dungern and Coca, Munch, med. Woch., 1907, p. 2317. 33 Von Liebermann. Biochem. Zeitschr., Vol. 4, 1907. 34Noguchi. Biochem. Zeitschr., Vol. 6, 1907. 35 Sachs and Kyes. Berl. kl. Woch., 2-4, 1903. 36 Hecker. Arb. auf dem konig. Inst. f. exp. Ther., Heft 3, 1907. 176 INFECTION AND RESISTANCE blood cells were thoroughly washed free of serum soaps did not have this hemolyzing action, and Friedemann and Sachs37 claimed that they were unable in any case to inactivate the hemolytic serum soap mixtures by heating to 56° C. These writers, as well as others, attribute Noguchi's results to the fact that the sera which he used to produce his "artificial complement," i. e., his serum soap mixtures, were heated to 50-51° C. only, a fact which would justify doubt of complete inactivation. Knaffl-Lenz38 has more recently carried out experiments on the same question. His results seem to show that the hemolytic action exerted by fatty acids or soaps is a phenomenon quite incomparable to true complement action, and that these hemoly- sins are heat stable, remaining unchanged by heating at 56° C. We have referred in a number of places to the analogy between alexins and ferments or enzymes. The chief objection to this conception formerly brought forward was based upon the fact that the comple- ment or alexin, unlike an enzyme, was used up during its reactions, and that a definite quantitative relationship existed between the alexin and the amount of cells or bacteria upon which it could act. Recent experiments by Kiss 39 seem to show that this quantitative re- lationship is not as strict and regular as was formerly supposed. lie showed that the action of complement depends very largely upon its concentration. For instance, to cite his work directly: "0.05 com- plement is sufficient to hemolyze completely a definite quantity of sensitized blood cells if the experiment is done in a total volume of 5 c. c. 0.02 c. c. of complement gives absolutely no hemolysis in a similar volume. When, however, the total volume is reduced to 2.5 c. c., then 0.02 c. c. of the complement begins to act, and it produces complete hemolysis if the total volume is reduced to 1.25." In fur- ther developing this observation he showed that, if sufficiently con- centrated, a very small amount of complement can act upon an ex- tremely large amount of red blood cells, an amount incomparably larger than those acted upon in more dilute solutions. These observa- tions would tend to strengthen considerably the conception of the fer- ment nature of alexins in general. Kiss' observations are furthermore in agreement with the inves- tigations of Liefmann and Cohn,40 whose work we have mentioned in the preceding chapter on Cytolysis. These writers assert that the fixation of complement during hemolysis is not due to its chemical union with the sensitized cells, but is due to fixation by the end products of the reaction ; in other words, by the stromata of the red cells and possibly by other substances given up by these cells. A further factor contributing to the disappearance of complement in 37 Friedemann and Sachs. Biochem. Zeitschr., Vol. 12, 1908. 38 Knaffl-Lenz. Biochem. Zeitschr., Vol. 20, 1909. 39 Kiss. Zeitschr. f. Imm., Vol. 3, 1909. 40 Liefmann and Cohn. Zeitschr. f. Imm., Vol. 8, 1911. FURTHER DEVELOPMENT OF KNOWLEDGE 177 such reactions is, they claim, its rapid deterioration at 37° to 40° C., when diluted. If they are right, these considerations also remove important objections to the conception of complement as a ferment. It is clear, therefore, that although we have gained much detailed information regarding the functional activity of the complement or alexin, and may assume, in a general way, that its action is similar to, if not identical with, that of an enzyme, we are nevertheless still very much in the dark concerning its chemical nature. The same thing may be said in regard to its physical characteristics. One method of investigating the physical properties of complement has been that of filtration. It may be remembered that one of Ehrlich and Morgenroth's 41 arguments in favor of the multiplicity of com- plement was the fact that, when goat serum was filtered through a Pukal candle, the complement which was active upon rabbit corpus- cles was retained, while that which acted upon guinea-pig cells passed through. Immune bodies or amboceptor always passed through. Vedder,42 in similar experiments upon bactericidal complements, claims to have been able, in the same way, to separate the comple- ments acting upon different bacteria. The problem has been more recently investigated by Muir and Browning.43 Their conclusions are briefly as follows: In the early stages of filtration through a Berkefeldt filter complement is often completely held back. After continued filtration it begins to pass through. If the complement is inactivated by the addition of hypertonic salt solution (5 per cent.), it passes through, and the filtrate can be reactivated by dilution to isotonicity. Sensitizer or amboceptor always passes through. Just how these experiments are to be* interpreted is a little obscure. The fact that the addition of salt renders the complement capable of passing through the filter would seem to indicate that its original inability to permeate did not depend upon the size of the molecule. On the other hand, it is also possible that the addition of salt to the complement may increase its dispersion in such a way that the indi- vidual particles are rendered smaller. This, however, is purely speculative, and we are at a loss for a fully satisfactory explanation of the results of Muir and Browning. We have repeated some of the experiments of Muir and Browning and, in substance, 'confirmed their results. It is our opinion that new filters remove complement by adsorption, just as this is accomplished when complement is shaken up with kaolin or other finely suspended material. That the addition of salts of various kinds in quantities greater than isotonicity (or more than the equivalent of 0.85=0.9 per cent. 41 Morgenroth and Ehrlich. Ehrlich's "Gesammelte Arbeiten," etc. 42 Vedder. Journ. Med. Ees., Vol. 9, 1903. 43 Muir and Browning. Journ. of Path, and Bact., Vol. 13, 1909. 178 INFECTION AND RESISTANCE NaCI)44 exerts a profound action upon the activity of complement is well known. Nolf 45 noted this in 1900, and the problem has been studied since that time by many investigators. Von Lingelsheim,46 who studied it in connection with his work on the refutation of the "osmotic" theories of immunity, showed that increasing the salt con- tents of serum (KNO3, Nad, K2HPO3, etc.) progressively dimin- ished its bactericidal power. Hektoen and Ruediger 47 also, after a very thorough study of this phenomenon, conclude that the action of the salts in such cases is exerted upon the alexin or complement and not upon the heat-stable sensitizers, and that it probably depends upon "physicochemical" causes. However, the manner in which such salt-inactivation is brought about is, to a great extent, obscure. There is no visible precipitation from serum after the addition of salts sufficient in quantity to weaken its action. Nothing is, as far as we can tell, removed from solution, and yet there is temporary inactivation which, at the same time, renders the complement fil- trable, facts from which we can only surmise some physical altera- tion. Inactivation of the complement also follows the removal of salts, but here the process is accompanied by a definite chemical change in that the serum globulins are precipitated. Studies of this process have led to important modifications in our conception of the nature of alexin, since they have shown that this body, formerly assumed to be single and homogeneous, may be sub- divided into at least two component parts by a number of experi- mental procedures. Ferrata was the first one to point this out as a consequence of investigations undertaken by him primarily with the purpose of determining the nature of the influence of salts upon hemolytic processes. Older studies of Buchner and Orthenberger 48 had shown that bactericidal action was inhibited when salts were removed from the medium, but the causes underlying such inhibi- tion had not been made clear. Ferrata 49 found, in the first place, that the absence of salts exerted no effect upon the mechanism of sensitization, but that amboceptor or sensitizer became attached to the cellular elements as readily when salts were absent as when the reagents were suspended in normal salt solution. It was a natural inference, therefore, that the failure of hemolysis, which he observed in •salt-free media (analogous to the similar experiences of Buchner 44 Alexin can be preserved in the refrigerator for long periods if hyper- tonic salt solution (15 to 25%) is added. It will again become active if iso- tonicity is restored with distilled water. 45 Nolf. Ann. Past., Vol. 14, 1900. 46 V. Lingelsheim. Zeitschr. f. Hyg., Vol. 37, 1901. 47 Hektoen and Ruediger. Journ. of Inf. Dis., Vol. 1, 1904. 48 Buchner and Orthenberger. Archiv f. Hyg., Vol. 10, 1C90. 49 Ferrata. Berl. kl. Woch., 1907, No. 13. FURTHER DEVELOPMENT OF KNOWLEDGE 179 in the case of bacteriolysis), must be attributed to failure of func- tionation on the part of the complement. On further investigation he obtained a very simple explanation. Ferrata removed the salts from his sera by dialyzing for twenty-four hours against distilled water. In this process, of course, there is a precipitation of the globulins while the water-soluble albumins remain in solution. The former may be redissolved in normal salt solution and the latter rendered isotonic by the addition of calculated amounts of concen- trated salt. In this way the original serum components are divided into two parts, neither of which, as Ferrata found, is alone capable of producing hemolysis of sensitized cells. In order to obtain the complementary action possessed by the original serum it is necessary to combine the two. This principle discovered by Ferrata is probably re- sponsible also for the results obtained by Sachs and Teruuchi,50 who likewise noted the destruc- tion of the complementary function in sera diluted with distilled water, but attributed this, in their publication, to the action of a comple- ment-destroying ferment, which is assumed to be active in salt-free media only. In his first experiments Ferrata reported that the precipitated globulin fraction was thermostable, the thermolability of complement CONCEPTION OP COM- being due entirely to the unprecipitated albu- S^A^I^R'ST min fraction. The work of Ferrata was soon SUGGESTED BY continued, however, by a number of other work- BRAND. ers, who confirmed the essential fact of the par- tition of the complement but modified and considerably extended the original observations. Brand 51 found that both fractions were equally thermolabile, and that the globulin sediment, after being redissolved in salt solution, could not be preserved in an active condition for more than a few hours. Preserved in distilled water or as sediment, it may retain its activity for several days, but dissolved in salt solution it becomes inactive within 3 to 4 hours, at room temperature. Michaelis and Skwirsky52 have since shown that the globulin frac- tion, thermolabile when free, is unaffected by a temperature of 56° C. after it has become attached to sensitized cells. Brand further studied the relationship of the two fractions to the sensitized cells and found that the globulin fraction may attach directly to such antigen-antibody complexes, but that the albumin fraction cannot be bound in this way unless the globulin fraction has been previously attached. For this reason he has referred to the former as the "end- 50 Sachs and Teruuchi. Berl kl Woch., 1907, Nos. 16, 17, and 19. 51 Brand. Berl. kl. Woch., 1907, No. 34. 52 Michaelis and Skwirsky. Zeitschr. f. Imm., Vol. 4, 1910. 180 INFECTION AND RESISTANCE piece" and the latter globulin sediment as the "mid-piece," assum- ing, on the basis of the conception of Ehrlich, that the globulin frac- tion serves to establish a link between the sensitized cell and the end-piece analogous to that formed by the "amboceptor" between the cell and the whole complement. It is possible, therefore, to treat sensitized cells with mid-piece in such a way that they are thereafter susceptible to hemolysis by the end-piece alone. Such cell-sensitizer- mid-piece combinations have been spoken of by Michaelis as 'fper- sensUized" cells. Tsurusaki53 confirmed the findings of Brand as to the thermo- lability of both "mid-piece" and "end-piece," but was unable to sep- arate the complement of normal hemolysins into the two components in the same way, since he found that hemolytic power was, in such cases, completely destroyed after twenty-four hours of dialysis. It seems to us not impossible that the natural deterioration of alexic power which takes place during such periods of time, at temperatures of from 16° to 20° C., may easily be held accountable for this, since the very feeble sensitization of cells in normal hemolysin complexes requires a correspondingly larger amount of alexin for activation. We have mentioned that the so-called "mid-piece" undergoes a rapid change when dissolved in salt solution and, after 3 or 4 hours, may lose its ability to induce hemolysis when added to sensitized cells together with end-piece. Although Hecker54 was able to con- firm this, he nevertheless showed that this fact does not imply a destruction of the mid-piece. For when such apparently inactive "mid-piece" was added separately to sensitized cells, and end-piece was subsequently allowed to act upon the complex, hemolysis re- sulted. This seems to show that the "mid-piece" undergoes a change on standing in salt solution which does not alter its ability to com- bine with the sensitized cells, but which subjects it to inhibition of such union when end-piece is present. It is also a peculiar fact, evi- dent in many of our own experiments, that when "mid-piece" and "end-piece" are first mixed and then added to sensitized cells the ef- fect in hemolysis is less powerful than when the "mid-piece" is added to the cells first, and the "end-piece" later. This effect is so instan- taneous that, if, in a series of experiments in which combinations of mid- and end-piece are used, the end-piece is run into the tubes con- taining the cells just before the mid-piece is added instead of the other way round, hemolysis is inhibited. In working with dialysis, also, wre have regularly had an experi- ence which may explain the difficulties which many other investiga- tors have had in such experiments. The globulin precipitate, whick 53 Tsurusaki. Siochem. Zeitschr., Vol. 10, 1908. 54 Hecker. Arb. a. d. konig. Inst. f. exp. Ther., Frankfurt a/M., Heft 3, 1907. See also Guggenheimer. Zeitschr. f. Imm., Vol. 8, 1911. FURTHER DEVELOPMENT OF KNOWLEDGE 181 fell out after dialysis of 24 hours or more, almost without exception, retained moderate or slight hemolytic properties, which could not be removed until the precipitate had been dissolved in salt solution and reprecipitated with distilled water two or three times. This would imply that a minute amount of the end-piece, carried down in pre- cipitation, must suffice to activate the mid-piece and would seem to point to the fact that in whole serum the two fractions are present as a complex and not separately. This question has been much dis- cussed and many facts have been brought out on both sides. Hecker showed that the combination of mid-piece with the sensitized cells can take place at a temperature of 0° C., while that of end-piece with the "persensitized" cells requires a considerably higher tem- perature. The bearing this fact may have upon similar earlier ex- periments of Ehrlich and Morgenroth upon the thermal conditions governing the union of amboceptor and complement with antigen is self-evident. In the present connection, however, the fact that the two fractions may be separately absorbed out of the serum by sensi- tized cells at 0° C. would suggest the probability of their being sep- arate in the whole blood. No crucial experiment has so far been possible, and there is not enough evidence on either side as yet to justify a definite opinion. However, the experiments of Michaelis and Skwirsky and later ones of Skwirsky alone have much indi- rect bearing on this question, though final interpretation is as yet impossible. Michaelis and Skwirsky,55 after determining that an acid reaction inhibits the hemolysis of sensitized blood cells, found that under such conditions "mid-piece" alone is bound, but that "end-piece" or the albumin fraction is left unbound. They recom- mend the use of strongly sensitized cells in an acid medium as a method of obtaining free "end-piece" from serum. Skwirsky56 subsequently found that durirg the ordinary Wasser- mann reaction the complex of syphilitic serum and antigen binds the mid-piece only. If the Wassermann reaction has been strongly positive; that is if there has been absolutely no hemolysis, and we remove the supernatant fluid by centrifugation, active end-piece can be demonstrated in it by the addition of persensitized cells. Bron- fenbrenner and RToguchi have also studied this phenomenon, but do not believe that Skwirsky's experiments prove that end-piece is free in such "fixation" supernatant fluids. These supernatant fluids, ac- cording to them, differ from all other "end-pieces" in that they are active upon persensitized sheep corpuscles only, but not^upon other cells. An explanation for this is lacking. There is much that is confusing in the facts so far revealed about the two component parts of the alexin. The most difficult fact to explain is the peculiar inactivation of the mid-piece in salt solution,, 55 Michaelis and Skwirsky. Zeitschr. f. Imm., Vol. 4, 1910. 56 Skwirsky. Zeitschr. f. Imm., Vol. 5, 1910. 182 INFECTION AND RESISTANCE which prevents its functionation in the simultaneous presence of end-piece, but does not seem to interfere with its ability to combine with the sensitized cells. As was to be expected, explanation for this has been sought by the Ehrlich school in changes of affinity. Sachs suggests that the mid-piece, by its preservation in salt solution, has lost its avidity for the sensitized cells and has gained in avidity for the end-piece, an alteration which therefore prevents its union with the cells. The same idea was suggested by Hecker himself. It is a little difficult to reconcile this explanation, however, with the fact that whole serum can be preserved and remain active in its comple- mentary function for a number of days, mid-piece and end-piece being present together, in a medium which, as far as salt contents are concerned, is isotonic with the salt solution in which mid-piece de- teriorates so rapidly wrhen alone. That there is, after all, much similarity between the alexins of different animals is evident from the fact that, as Marks and others have shown, the end-piece of one animal may activate the mid-piece of another species. It appears also from experiments like those of Ritz and Sachs 5T that an animal may possess a mid-piece for certain sensitized cell complexes without possessing a corresponding end- piece. Thus they found that the serum of mice contained a mid- piece but not an end-piece for sensitized guinea pig corpuscles. Much that has been found out about the so-called globulin portion, moreover, tends to engender doubt as to the wisdom of applying to these complement fractions the terms "mid-piece" and "end-piece," an objection which is based upon reasons similar to those which pre- vent Bordet from accepting the term amboceptor. For so little is actually known concerning the mechanism of complement functiona- tion, that it seems unwise to establish on a firm basis a preconceived idea of the mechanism by adapting the terminology to a theory. The most confusing feature of the problem lies in the surprising quantita- tive relations which seem to exist in the reactions of the two frac- tions. Thus Liefmann and Cohn58 claim that in the presence of moderately sensitized cells no measurable amount of the so-called mid-piece or globulin fraction is bound, that is, removed from solu- tion ; and yet, when both fractions are added to such cells, rapid and complete hemolysis results. In the presence of heavily sensitized cells (20 to 50 units) a small quantity only is removed. Nevertheless this fraction has had a demonstrable effect on the cells, since it has rendered them amenable to the action of the albumin fraction. In all such experiments, therefore, as Liefmann justly points out, the de- gree of sensitization must be taken into consideration before conclu- sions are formulated. It is curious also that a slight excess of the globulin fraction may prevent complement action completely. In 57 Ritz and Sachs. Zeitschr. f. Imm., Vol. 14, 1912. 58 Liefmann and Cohn. Zeitschr. f. Imm., Vol. 7, 1910. FURTHER DEVELOPMENT OF KNOWLEDGE 183 experiments cited by Marks 59 it appears that the most ineffective complement is obtained when "mid-piece" and "end-piece" are added to the sensitized cells in proportions of 1 to 1. If the proportion of "mid-piece" is increased two or threefold over that of "end-piece," hemolysis is inhibited. This, however, is true only when the two fractions are simultaneously added to the sensitized cells. When the sensitized cells are exposed to the excessive quantity of the "mid- piece" separately, and "end-piece" added later, the effect is one of stronger hemolysis than when smaller amounts are used. It is thus seen that the relations between the complement fractions in hemolysis are very involved. All that we can be sure of is that there are at least two separable parts, that one of these acts directly upon the sensitized cells, forming a so-called persensitized complex and rendering them amenable to the subsequent action of the unprecipitated albumin fraction. The many difficulties encountered in the interpretation of the confusing phenomena observed in connection with this problem have, very naturally, led to a corresponding multiplicity of opinion. Most observers at present incline to the opinion that the globulin and albumin portions of fresh serum, separated by Ferrata's or any other of several common methods, represent actually two complement frac- tions. This is not, however, accepted by all workers. Bronfenbren- ner and ^oguchi 60 believe that the entire active complement is con- tained in the albumin fraction or so-called "end-piece." They hold that "complement-splitting7 ' by dialysis or other methods is an inacti- vation of end-piece by change of reaction. In their experiments they were able to restore the functional activity of end-piece by the adjust- ment of reaction, either with acid or alkali, respectively, or by the addition of amphoteric substances. The mid-piece activates, they be- lieve, by reason of its amphoteric nature, and consequently adjusts any excessive acidity or alkalinity of the medium. They were able to substitute for mid-piece indifferent amphoteric substances such as alanin. Liefmann 61 has been unable to confirm the experiments of Bronfenbrenner and Noguchi, and believes that their results were caused by incomplete splitting of the complement. Incidental to a study of normal opsonins the writer has also repeated the experi- ments of Bronfenbrenner without being able to confirm them.62 The method of Ferrata for the separation of the two parts of the complement is successful only if dialysis is very thorough and suffi- ciently prolonged to lead to complete precipitation of the globulins. INeufeld and Haendel 63 have had difficulty in thus separating the 59 Marks. Zeitschr. f. Imm., Vols. 8 and 11, 1911. 60 Bronfenbrenner and Noguchi. Jour, of Exp. Med., Vol. 15, 1912. 61 Liefmann. Weichhardt's Jahresbericht, Vol. 8, 1912. 62 Zinsser and Cary. Journ. of Exp. Med., Vol. 19, 1914. 63 Neufeld and Haendel. Arb. a. d. kais. Gesund., 1908. 184 INFECTION AND RESISTANCE fractions, and the writer has noticed similar failures but has always been able to obtain eventual separation by sufficient prolongation of the dialysis. Because of the occasional difficulties and because of the time-consuming and inconvenient nature of the method other means of separation have been devised. The one used with success by many workers has been that introduced by Sachs and Altmann,64 namely, precipitation of the sera with weak hydrochloric acid ^^ to T^. Liefmann has separated the components by precipitation of the globulins by the introduction of CO2. In carrying out this method, Fraenkel 65 has found it advantageous to dilute the serum ten times with distilled water, then allowing the CO2 to flow in at low temperatures. It is likely that any of the usual methods of globulin separation will serve for complement partition. The salt- ing out methods are, however, extremely inconvenient because of the prolonged dialysis subsequently necessary to remove the salts. The inactivation of complement or alexin by the addition of salts or by splitting is, very apparently, a temporary inactivation in which prompt restitution can be practiced by bringing back original conditions either by dilution to isotonicity or by reconstruction of the divided substance, respectively. Heating to 56° C., the simplest and most commonly employed method of inactivations was, until of late, regarded as an irreversible process, the complement being irre- trievably destroyed in the procedure. Gramenitski 66 has recently carried out experiments which seem to show that this opinion is erroneous. His experiments were suggested by the fact, observed by Bach and Chodat,67 that certain oxydases and diastases may spon- taneously regain some of their activity after inactivation by heat. His work with complement indicated a similar gradual return to an active condition after moderate heating. The great theoretical im- portance of this observation will justify our insertion of one of Gramenitski's protocols. Experiment 1. Complement 10 times diluted was heated to 56° C. for 7 minutes. It was then tested against sensitized beef blood at varying intervals as follows: Time after heating at which test was made Quantity of hemoglobin gone into solution after % 10 min. % 20 min. % 30 min. % 40 min. Immediately after heating \^/2 hour 0 0 20 10 20 30 70 40 40 60 80 70 70 80 100 24 hours 48 hours 64 Sachs and Altmann. Cited from Sachs in "Kolle u. Wassermann Hand- buch," Vol. 2, p. 877. 65 Fraenkel. Zeitschr. /. Imm., I, Vol. 8, 1911. 66 Gramenitski. Biochem. Zeits., Vol. 38, 1912. 67 Bach and Chodat. Cited from Gramenitski, loc. cit., p. 511. FURTHER DEVELOPMENT OF KNOWLEDGE 185 In other experiments in which heating was more prolonged a similar regeneration was observed, though not as pronounced as in the one cited above. The largest amount of restored complement seemed to be present after about 24 hours. After this gradual de- terioration again ensued. It is quite impossible to offer an adequate explanation for this at the present time. Gramenitski 68 acknowl- edges this, but permits himself certain speculations which we repeat in nearly his own language, since there is much in them which seems to us reasonable. The complement, as indeed all other active serum constituents, must be looked upon as colloidal in nature. When heat is applied to such substances alterations occur which gradually lead* to coagulation. As this occurs there is an aggregation of particles and a consequent diminution of surface tension. This last point has been experimentally demonstrated by Traube,69 who has regularly found a fall of surface tension as serum was heated to 56° C. And of greatest interest in this connection is the further determination by Traube that a gradual restoration of the surface tension takes place as the serum is allowed to stand. It is not inconceivable, therefore, that the inactivation of complement by heat may depend upon an alteration of its colloidal state, i. e., an aggregation of the particles, which, if not carried too far, may be reversible and followed by a gradual dispersion as the serum is kept 24 hours. On the same grounds the gradual deterioration of complement on standing may be compared to the slow settling out of colloidal suspensions which eventually results in spontaneous precipitation, a process which occurs not only in chemically well-defined colloids, but is often ob- served in sera. Bechold has referred to this as "das Altern Kol- loidaler Losungen." Of great interest, furthermore, in connection with the physical properties of complement is the discovery made by Jacoby and Schiitze 70 that complement can be inactivated by shaking. This astonishing observation has been confirmed by Zeissler,71 Noguchi and Bronfenbrenner,72 Ritz,73 and others. It appears, according to these observers, that guinea pig serum, when subjected to active shaking, can eventually be robbed thereby of its activating prop- erties. The success of such experiments depends somewhat upon the concentration of the serum, and is best observed in a dilution of 1 part, to 10 parts of salt solution. Under such conditions complete inactivation may be observed within 20 to 25 minutes. Between the inactivation of complement by heat and that which results from *8 Gramenitski. Loc. cit., p. 504. 69 Traube. Zeitschr. f. Imm., Vol. 9, 1911, and Biochem. Zeitschr., 1908. 70 Jacoby and Schiitze. Zeitschr. f. 1mm., Vol. 4? 1910. 71 Zeissler. Berl. kl. Woch., No. 52, 1909. 72 Noguchi and Bronfenbrenner. Journ. of Exp. Med., Vol. 13, 1911. 73 Ritz. Zeitschr. f. Imm., Vol. 15, 1912. 186 INFECTION AND RESISTANCE shaking, there are certain similarities which seem to strengthen the opinion regarding the nature of heat inactivation which we have cited above. For it has been variously shown that prolonged shaking of protein solutions, like heating, gradually leads to coagulation. It would be important to determine whether or not the inactivation by shaking, like that produced by heat, is accompanied by a fall of sur- face tension. ALEXIN OR COMPLEMENT FIXATION The controversy regarding the multiplicity of alexin and the existence of a "complementophile group'7 cannot, of course, be re- garded as closed, however much we may lean toward the acceptation of Bordet's point of view, since German experimenters of eminence still adhere to the Ehrlich interpretations. Moreover, it is, of course, extremely difficult to disprove such an assumption as that of the "polyceptor" conception of the complementophile group. However, we may safely assert that the functional unity of complement (and, after all, that is all that Bordet has maintained) is being upheld by the constantly increasing evidence in its favor which is being fur- nished by the practical and experimental application of the phe- nomenon of "alexin fixation" described, in 1901, by Bordet and Gengou.74 It will be well to bear in mind that this phenomenon should be strictly distinguished from the so-called "complement de- viation" ("Ablenkung"), described by Neisser and Wechsberg. The latter was advanced as an explanation of the inactivity of bacteri- cidal sera when used in too great concentration, as described in an- other place (p. 160) (Neisser and Wechsberg phenomenon), and has been variously utilized as support for the assertion that alexin can unite with unattached sensitizer. It is regarded by most observers, moreover, as untenable in the light of later investigation. In spite of this, the term "Komplement-Ablenkung" has been employed by a number of German writers (see Citron, Vol. 2, "Kraus und Levaditi Handbuch") as synonymous with "fixation" in the sense of Bordet and Gengou. The phenomenon of Bordet and Gengou, briefly described, is nothing more than an experimental utilization of the fact which we have discussed at length, that alexin is fixed by antigen and antibody after union, but by neither alone. The condition, as observed by them, may be best described by submitting the protocol of the first experiment detailed in their communication : An emulsion of a 24-hour slant of plague bacilli was used as 74 Bordet and Gengou. Ann. de I'Inst. Past., 1901, Vol. 15, p. 289. FURTHER DEVELOPMENT OF KNOWLEDGE 187 antigen, heated antiplague horse serum represented the antibody, and fresh guinea pig serum was usej, as alexin. A series of tubes was then prepared as follows : . ^ Hague bacrlli + inactivated antipl 2. Alexin -j- plague bacilli + inactivated normallibrse serum. 3. Alexin + inactivated antiplague serum. 4. Alexin + inactivated normal horse serum. 5. Plague bacilli + inactivated antiplague serum. > 6. Plague bacilli and normal horse serum. These mixtures were left together for 5 hours and, at the end of this time, sensitized rabbit corpuscles were added to each tube. The result showed hemolysis in all the tubes except "1," in which there were plague bacilli, antiplague serum, and alexin, and in tubes 5 and 6, which had contained no alexin from the beginning.75 It was plain, therefore, that the bacilli when specifically sensi- tized had become capable of absorbing alexin and preventing its sub- sequent action upon the sensitized erythrocytes. That the occurrence was not exceptional was shown by the fact that, in the same series, similar results were obtained with anthrax, typhoid, and proteus bacilli, and their respective antisera. Schematized in accordance with the conceptions of Ehrlich our diagram would be as follows : +- COMPLEMENT Oft AL&UN SPECIFIC ANT/BODY — » (P&SENT O&NOT?} ANTIGEN _^ (BACTERIA ETC.) HAEMOLYTIC ANTIBODY r- REDBLOOD, CELLS I I , COMPLEMENT FIXATION SCHEMATIZED ACCORDING TO EHRLICH 's VIEWS. If the antibody in I is present then complement is fixed by the antigen-antibody complex, and is no longer free to act upon the hemolytic complex II. In the same way antigen I could be determined if a known antibody I were used. For, in the absence of either of these parts of the complex I complement would remain unfixed and free to act on complex II. We represent the phenomenon graphically in the symbols of Ehr- lich merely because they facilitate clearness of exposition. In the presence of both parts of Complex I the alexin is held and 75 We will see later that unsensitized bacteria in emulsion will non-specifi- cally fix small amounts of complement. 188 INFECTION AND RESISTANCE is no longer available for Complex II. If either of the reacting parts, antigen or sensitizer, of Complex I are lacking the alexin is left unfixed and free to react with Complex II. With this technique Bordet and Geiigou were able to demonstrate, by indirect experiment, the presence of specific sensitizers in the sera of animals immunized with various bacteria, a fact which was, of course, surmised but had been amenable to proof heretofore only in the case of bacteria like the spirillum of cholera in which lysis under the influence of immune serum and alexin could be directly observed under the microscope. The practical possibilities of their method were, of course, immediately apparent. By the use of a known antigen specific sensitizers can be demonstrated in this way, and, vice versa, in the presence of a known antibody, the method will serve to identify the nature of a doubtful antigen. Thus bac- terial differentiation can be carried out by adding to the suspected bacteria, in emulsion, a small quantity of a known antiserum and alexin, and determining whether or not the alexin has become fixed. And, conversely, Bordet and Gengou 7G have more recently utilized the method in support of their claim of the specific etiological im- portance of the bacillus isolated by them from whooping cough, by showing that the serum of children suffering from this disease formed a specific alexin-fixing complex when treated with the ba- cillus. The phenomenon of Bordet and Gengou thus found rapid prac- tical application in the diagnosis of a number of infectious diseases, and has, of course, attained great clinical importance in the diag- nosis of syphilis in the form of the "Wassermann" reaction and its many modifications. Before discussing these practical features in greater detail, however, it will be useful to discuss more particularly the many important theoretical considerations which have followed in the train of the complement-fixation phenomena. A year after the publication of Bordet and Gengou's paper Gengou 77 made another fundamentally important observation by showing that complement or alexin fixation was not limited to the complexes of cellular antigens and their antibodies, but that the sera of animals immunized with dissolved proteins (animal sera, etc.), when brought together with their specific antigens, likewise formed combinations which fixed alexin. Thus egg-white or dog serum, brought together with "anti-egg-white" or "anti-dog" rabbit serum, respectively, strongly fixed alexin, whereas neither the antigenic sub- stances nor the antisera exerted such fixation alone. The interpre- tation put upon this by Gengou was the following : "In sera obtained by injecting rabbits with large doses of cow's milk, etc., there are, in addition to the precipitins of Bordet and Tschistovitch, substances 76 Bordet and Gengou. Ann. de I'Inst. Past., 1906. 77 Gengou. Ann de I'Inst. Past., 16, 1902. FURTHER DEVELOPMENT OF KNOWLEDGE 189 analogous to the sensitizers described by Bordet in bacteriolytic and hemolytic sera, and later found in the majority of antimicrobial sera." The important point in this interpretation is that Gengou conceived the existence of antiprotein sensitizers, in addition to the precipitins, formed as a response to immunization with amorphous protein. Moreschi 78 soon confirmed Gengou's experimental deter- minations, and Neisser and Sachs79 took the further logical step of applying this knowledge to the determination of proteins for foren- sic purposes. This, too, we will further discuss when we speak of the practical features of these phenomena. It thus appears that the fixation of alexin is a generalized property of all mixtures in which an antigen is brought into contact with its specific antibody, whether the antigen is in the form of the whole bacterial or other cell, or in that of a dissolved protein, animal serum, or egg-white, etc. The observation of Gengou, though for a time insufficiently val- ued, has had a profound influence upon the subsequent under- standing of serum reactions. The fundamental importance of this work was not fully recognized until his studies had found logical continuation in the investigations of Gay80 and in those of Moreschi. Moreschi 81 studied the antihemolytic properties possessed by the serum of a rabbit which had been treated with normal goat serum. He found that such a serum had distinct anticomplementary powers when it was added to a hemolytic system of ox blood sensitizer (ob- tained against ox blood from rabbits), and goat complement. With such a hemolytic system, however, there was anticomplementary ac- tion only against goat complement and not against rabbit or guinea pig complement. If, however, he used a hemolytic system in which the amboceptor or hemolytic sensitizer employed was one obtained from a goat, the serum was anticomplementary for all complements which were used. Moreschi concluded from this that the apparent anticomplementary action of the serum could not be interpreted as the action of a specific anticomplement in the sense of Ehrlich, but that it resulted from the reaction which took place as the consequence of union of the antibody in the anti-goat rabbit serum and goat pro- tein, which was introduced into the tubes, in the first case with the complement, and in the second with the amboceptor. He proved his contention by obtaining similar universal anticomplementary action when he added a little normal goat serum to the tubes set up as above described. It is plain, therefore, that anticomplementary action can be explained in observed cases by the simple consideration of the phe- nomenon of Gengou. Similar findings were later recorded by Muir 78 Moreschi. Berl. kl. Woch., No. 37, 1905. 79 Neisser and Sachs. Berl kl Woch., No. 44, 1905. 80 Gay. Centralbl f. Bakt. I Ori 300 B. typhi 4,000 B. coli" 31" 0 B. typhi 0 £.coB"31"0 In the preceding paragraphs, however, we have seen that im- munization with a single organism, say B. typhosus, will induce the formation of agglutinins, not only for this species, but also of para or minor agglutinins for biologically similar strains as well. In such cases, as Castellani showed, absorption of the serum with the organism used for immunization takes out, not only the major ag- glutinins, but rather all of the agglutinins, major and minor. Con- versely, however, absorption of such a serum with the species ag- glutinated by the minor agglutinins takes out these antibodies only, leaving the major substances intact. These relations are well illus- 42 Pfeiffer. Quoted from Paltauf in "Kolle u. Wassermann Handbuch," Vol. 4. 43 Castellani. Zeitschr. /. Hyg., Vol. 40, 1902. THE PHENOMENON OF AGGLUTINATION 233 trated by the two following protocols, also taken from Castellani's paper: ' Serum of rabbit No. 1 immunized with B. typhi only Agglutination titre of serum Titre after absorption with B. typhi B typhi 5,000 B. coli 600 B. typhi 0 B. coli 0 Serum of rabbit No. 7 immunized with B. typhi only Agglutination titre of serum After absorption with B. typhi After absorption with B. coli B. B. typhi 10,000 (heavy clumps) coli 800 B. typhi 0 B. coli 0 B. typhi 10,000 (small clumps) B. coli 0 Note: All of these protocols are taken from Castellani's communication, loc. cit. These facts, variously confirmed, tend to corroborate the concep- tion of the production of major and minor agglutinins outlined above. It is of practical and theoretical importance to mention that complete absorption of specific agglutinin by a single exposure to homologous bacteria, however thickly emulsified, is not possible. It is always necessary to absorb repeatedly, and even then a minute trace of agglutinin may eventually remain. Eisenberg and Yolk,44, 45 who have studied these conditions particularly, attribute this to the nature of the union of agglutinogen with agglutinin, which they conceive as following the laws of mass action — this ac- counting for the persistence of a small "rest" of free agglutinin, even after repeated absorption by partial dissociation. The prin- ciple involved here is identical with that discussed in connection with antigen-antibody union in general. It is not only in the sera of immunized animals, however, that agglutinins are found. Just as the other antibodies, antitoxins, and bactericidal sensitizers may be found in the blood of normal animals, so agglutinins for various bacteria may be normally present. These normal agglutinins do not in any respect, further than that of quan- tity, differ from the immune agglutinins and follow the same laws of specificity which have been described for the latter. It has been shown a number of times that such normal agglutinins are not pres- 44 Eisenberg and Volk. Zeitschr. f. Hyg., Vol. 40, 1902. 45 Eisenberg. Centralbl f. Bakt., Vol. 34, 1903. 234 INFECTION AND RESISTANCE ent in the new-born animal, but are acquired later in life, possibly because of the absorption of bacterial products from the intestinal canal. It has been variously shown,46 too, that living bacteria them- selves may enter the lymphatics and the portal circulation from the intestine during apparently perfect health of the individual. This subject is of interest, not only in connection with the ag- glutinins, but has bearing upon the existence of normal antibodies in general. Ruffer,47 who has studied particularly the penetration of leukocytes and bacteria through the intestinal mucosa, demon- strated micro-organisms in the sub-mucous lymph nodes of normal rabbits, and Ribbert 48 and Bizzozero 49 have shown the presence of bacteria in apparently normal mesenteric lymph nodes. Adami and Nichols even claim that during health a certain number of liv- ing bacteria enter the portal circulation from the intestine, and from here may get into the systemic circulation, and are ordinarily de- stroyed by either leukocytes, liver lymphatic organs, or the kidneys. It is thus not surprising that normal agglutinins should occur, and that they should be qualitatively identical with the so-called "immune" agglutinins, since they probably arise by a sort of spon- taneous immunization through the intestinal canal. From the in- vestigations of Ford especially we may conclude that the immune agglutinin may be regarded as merely a quantitative increase of the normal antibody, if this has been present before immunization. Ford50 found that when an animal is treated with an agglutinating serum an anti-agglutinin may be obtained which neutralizes the action, not only of immune, but also of homologous, normal ag- glutinin. An interpretation of the process of agglutination, according to the theory of Ehrlich, conceives it as a chemical union of agglutinin and bacteria (agglutinogen). The agglutinin is regarded as con- sisting of two atom complexes, one the "haptophore," having af- finity for the bacterial protein, and concerned with the union, the other the "ergophore" or "zymophore," by means of which the actual agglutination is brought about after the union has taken place. Un- like the antibody concerned in the processes of hemolysis or bac- teriolysis, the agglutinins are not dependent in their action upon the cooperation of alexin, and the agglutination power of a serum is therefore not 'destroyed by inactivation or heating to 56° C., as is the case with the former. Although the accurate point of thermal destruction varies with different agglutinins (the agglutinins for the Bacillus pestis and a few other bacilli are said to be destroyed at 46 Adami. Jour. Am. Med. Assoc., Dec., 1899. 47 Ruffer. Brit. Med. Journal, 2, 1890. 48 Ribbert. Deutsche med. Woch., 1885. 49 Bizzozero. Centralbl. f. d. Med. Wiss., Vol. 23, 1885, p. 49. Quoted from Adami. 50 Ford. Zeitschr. f. Hyg., Vol. 40, 1902. THE PHENOMENON OF AGGLUTINATION 235 56° C.), as a rule agglutinins will not disappear from serum until the temperature is raised to between 70° and 80° C. Once de- stroyed, however, no reactivation takes place upon the addition of fresh normal serum. Ehrlich, for this reason, has conceived the structure of agglutinins as "Ilapiines of the Second Order," in which he supposes that the zymophore and the haptophore groups are inseparably connected, and in which we could assume an alteration of the less stable zymophore group without interference with the functional integrity of the haptophore group. Such an altered ag- glutinin could be spoken of as "agglutinoid," and this could become united with a bacterial cell without inducing agglutination, but, by its union, prevent subsequent combina- tion of the cell with unaltered agglu- tinin. This process of "agglutinoid Yerstopfung" has been held respon- sible for the failure of agglutination when bacteria that have been in con- tact with heated serum are subse- quently exposed to the action of ac- tively agglutinating serum. It is •^ i ;? t .• -i i • i DIAGRAMMATIC REPRESENTATION OF assumed that the agglutmoids which EHRLICH 's CONCEPTION OF THE were present in the heated serum STRUCTURE OF AGGLUTININ. have occupied the bacterial receptors and have thereby prevented the union of these with the agglutinins later added. The work of Eisenberg and Yolk 51 has gone very thoroughly into these conditions and forms the strongest bulwark of this point of view. These workers showed that bacteria thus exposed have not only become less sensitive to agglutinins, but have, at the same time, lost much of their power to absorb agglutinins when compared with normal bacteria. The same loss of agglutinating power which is observable in heated agglutinating serum is evident to a lesser extent in serum wrhich has been preserved at room temperature. Eisenberg and Yolk have shown that such serum, in addition to a loss of its total agglutinin content, loses the power to agglutinate in high concentrations. Thus a serum which has been preserved in this way will no longer agglutinate bacteria in concentrations of 1 to 20, 1 to 40, or even 1 to 100, but will agglutinate as before in higher dilutions. This is taken to mean that the agglutinoids formed during the period of standing possess a higher affinity for the bac- terial antigen than do the true unaltered agglutinins. Since these so-called "proagglutinoids" are relatively small in amount, their action is masked when considerable dilution has sufficiently di- minished their quantity, in proportion to the more plentiful un- 51 Eisenberg and Volk. Zeitschr. f. Hyg., Vol. 40, 1902. 236 INFECTION AND RESISTANCE altered agglutinins. In support of this assumption it has been further shown that sera which have been rendered inhibitory by either of the methods named can be deprived of their inhibiting characteristics by absorption with homologous bacteria. Together these observations constitute a strong argument in favor of the ag- glutinoid theory. In practical experience the existence of such an inhibition zone is of great importance, since freshly taken sera will occasionally show this failure of agglutination in concentration, while strong agglutination follows when the dilution is increased. In clinical tests, as in the Widal reaction for the diagnosis of typhoid fever, we not infrequently encounter sera which will give no agglutination in dilutions of 1 to 20 and even 1 to 40, and the reaction would therefore be regarded as negative unless the possibility of the pro- agglutinoid zone were recognized and further dilutions carried out. While there is no question of the accuracy of the experimental data cited in the preceding paragraphs, the interpretation of the phe- nomena on the basis of Ehrlich's haptine conception has not been universally accepted. The fact that the agglutinins lose their agglutinating power after heating, while retaining their power to prevent the subsequent agglu- tination of the bacteria, may be more simply explained in analogy with the observation of Forges on the influence of heated serum upon the agglutination of mastic suspensions. He found that unheated serum will flocculate such suspensions, while heated serum of the I same concentration will prevent the flocculation, acting probably as J a protective colloid (see chapter on Colloids). In the same way the heated agglutinating serum may prevent subsequent flocculation by a similar protective action. We suggest this as a possible explanation of the proagglutinoid phenomenon, although of course it is a mere conjecture as opposed to the painstaking experiments of Eisenberg and Yolk. It has the advantages of simplicity, but does not, it is true, explain the apparent specific absorption of the agglutiniii- inhibiting properties out of heated sera with homologous bacteria, as claimed by these authors as well as Kraus and Joachim. The simi- larity of the proagglutinoid phenomenon to the inhibition zones oc- curring in the flocculation of colloids will be referred to in a subse- quent paragraph. In describing the investigations which led to the discovery of the mechanism of the lytic phenomenon, in the chapter on Cytolysis, we mentioned that Bordet and others had noticed the frequent ag- glutination of red blood cells in the sera of animals treated with such cells after the hemolytic property had been destroyed by heat- ing to 56° C. Such hemagglutination is a phenomenon entirely analogous to the agglutination of bacteria by serum, and hemag- glutinins regularly result when an animal is systematically treated THE PHENOMENON OF AGGLUTINATION 237 with the red blood cells of another species. Like the bacterial ag- glutinins, the hemagglutinins are relatively thermostable and are best observed, therefore, after the sera are inactivated. Otherwise rapid hemolysis will often obscure the agglutination. Like other agglutinins the hemagglutinins thus produced are specific, acting only upon that variety of cells which are used in their production. Moreover, certain sera may normally contain hemagglutinins for they"' blood cells of animals of another species. An illustration of this is the hemolytic and hemagglutinating property of normal goat serum for rabbit cells — but there are many other examples which might be cited. Such normal hemolytic and hemagglutinating properties for the cells of other animals usually render the sera toxic for these ani- mals, and some observers have attributed the toxicity to this agglu- tinating action, believing that the clumped erythrocytes form emboli around which clotting is initiated. The writer's own investigations, however, seem to show that this is not the case, since the toxicity of such sera is completely removed after they have been heated, in spite of the fact that the hemagglutinative property remains unchanged. In discussing hemolysins, also, we called attention to the observa- tion that the sera of animals may develop the property of hemolyzing blood cells of other individuals of the same species when immunized with such cells, and that on occasion such isolysins may be normally present. Analogous to iso-lysins, iso-agglutinins also have been observed. They were described first in human blood in 1900, independently, by Landsteiner,52 and by Shattock. As the result of the work of a number of men, in particular that of Landsteiner, of Ascoli,53 and of Descatello and Sturlii,54 it became evident that all human blood fell into four sharply separated and, for the individual, permanent groups, according to the way in which they interagglutinate. The members of the first group have blood cells which are not agglutinated by the serum of any human blood. The serum of the members of this group agglutinates the blood cells of persons belonging to any of the other three groups. This group constitutes between 40 to 50 per cent, of all human beings. Members of the second group have blood serum which agglutinates the cells of persons belonging to the third and fourth groups ; while the cells of the second group are ag- glutinable by the serum of the first and third groups. The third is the reciprocal of the second, and the serum of the third group ag- glutinates the cells of the second and fourth groups ; while the cells of the third group are agglutinated only by the serum of the first and second groups. The fourth group (which is very rare) is the recipro- 52 Landsteiner and Richter. Zeitschr. f. Med., 3, 1902. 53Ascoli. Munch, med. Woch., 1901. 54 Descatello and Sturlii. Munch, med. Woch., 1902, 49, p. 1090. 238 INFECTION AND RESISTANCE TABLE FOR ISO-AGGLUTININS Sera i I II III IV 1 2 3 4 5 6 7 8 9 10 r 1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 I\ 3 0 0 0 0 0 0 0 0 0 0 • 4 0 0 0 0 0 0 0 0 0 0 \ 5 + + + + 0 0 0 + + 0 6 + + + + 0 0 0 + + 0 I 7 + + + + 0 0 0 + + 0 ml 8 + + + + + + + 0 0 0 1 9 + + + + + + + 0 0 0 ivj 10 t + + + + + + + + 0 * For this table as well as for much direct information concerning the iso-agglutinins and iso- lysin we are indebted to Dr. Ottenberg of this laboratory. cal of the first, the serum having no agglutinin, the cells being ag- glutinated by the serum of any other group. (See table.) It is evident from examining this grouping that the phenomena can be explained (as Landsteiner has suggested) if it is assumed that there are two agglutinins (a and ft) and two corresponding agglutinogens present in the red cells (A and B). The blood of the first group pos- sesses both agglutinins, but no agglutinogens, the blood of the second group possesses agglutinin a, agglutinogen B, the blood of the third group possesses agglutinin ft, agglutinogen A, the blood of the fourth group possesses no agglutinin but both agglutinogens. These agglutinins are present in weak dilution only, being gen- erally active in dilutions only of 1-15 to 1-30. They are separately absorbed from the serum by the suitable red cells (Hektoen).55 Ot- 55 Hektoen. Jour, of Inf. Dis., 1907, p. 297. THE PHENOMENON OF AGGLUTINATION 239 tenberg noticed that they were inherited, and this was also shown in 1908 and in 1910 by von Dungern and Hirschfeld,56 who further found that this inheritance followed the Mendelian law strictly. The agglutinogens are the unit characters. The agglutinogens ap- parently are present at an earlier embryonic stage than the ag- glutinins. On account of their hereditary nature and permanence for the individual the iso-agglutinins may possibly be of medicolegal value. They may also be of some practical importance in selecting donors for blood transfusion, as phagocytosis of red blood cells in the circulation after transfusion, first described by Hopkins, was proved by Ottenberg to occur only when the patient's serum was agglutina- tive toward the donor's red cells, and several such transfusions have had fatal results. Similar iso-agglutinins have been observed in the blood of lower animals, in horses (Klein,57 1902) ; rabbits (Boycott and Douglas,58 1910) ; cats (Ingebrigtsen) ; dogs, rats, and steers (Ottenberg).59 The iso-agglutinins have been developed in dogs (von Dungern and Hirschfeld).60 In most of the lower animals they have occurred with peculiar irregularity, indicating probably the presence of, not two, but of a larger number of agglutinins and agglutinogens. In steers, however, they fall into simple groups, indicating the presence of only one agglutinin and agglutinogen. In many animals the agglutinins are entirely latent, and the biochemical differences represented by the agglutinogens are present in the red cells, and the correct agglutinin is developed by the animal only when it is immunized with blood whose cells contain agglutinogen not present in the animal's own blood cells/ The fundamental principle underlying all the Ehrlich hypotheses is the conception that these reactions take place as do purely chemical reactions, following the law of multiple proportions. Such reasoning has often necessitated the assumption of differences of affinity which, critically examined, are really ex post facto explanations, forcing the phenomena to conform with the theory. As a matter of fact, the bodies which participate in the antibody-antigen reactions are probably of the nature of the substances which are spoken of as col- loids, and it is therefore more than likely that the quantitative and other relations governing the union of these reagents should be anal- ogous to those governing colloidal reactions in general. The reaction of agglutination, like that of precipitation, has lent itself particularly to the study of the principles of the union from this point of view, and the first and fundamental progress made in this direction is found in the work of Bordet. 56 Von Dungern and Hirschfeld. Zeitschr. f. Imm., 4, 1909-1910, p. 53; also ibid., 1910, p. 284. "Klein. Wien. kl Woch., 1902, p. 413. 58 Boycott and Douglas. Jour, of Path, and Bact., Jan., 1910. 'a Epstein and GHtenber^. Tr. N. Y. Path. Soc., 1908. 60 Von Dungern and Hirschfeld. Zeitschr. f. Imm., 1909, 1910, p. 531. 240 INFECTION AND RESISTANCE Bordet 61 compared the formation of precipitates in bacterial emulsions to the precipitation of such inorganic emulsions as clay in distilled water, and noted that the precipitation of homogeneous emulsions of such substances is "often controlled by such insignifi- cant causes as the presence of salts." Applying this analogy to the agglutination of bacteria, he performed the following experiment: Cholera spirilla, emulsified in salt solution, were treated with homol- ogous immune serum and, after agglutination had taken place, the bacteria were thrown to the bottom by centrifugation and divided into two parts. One part was again suspended in salt solution, and the other was washed, and then suspended in distilled water. The bac- teria in the tube of salt solution rapidly agglutinated, while those in the distilled water, after thorough shaking, remained indefinitely suspended in an even emulsion. If, however, to these unagglutinated bacteria a small amount of pure sodium chlorid was added agglutina- tion occurred. The conclusions that can justly be drawn from this experiment are, first, that the bacteria could not agglutinate, even though they had been bound to agglutinin, when salt was removed from the en- vironment, and, second, that the addition of salt to such emulsions brought about immediate agglutination. The same principle can be demonstrated in other ways. If, for instance, a bacterial emulsion is rendered free of salt by dialysis, and this is added to an aggluti- nating serum similarly dialyzed, no agglutination occurs. The sus- pension may remain evenly clouded indefinitely unless salt is added. As soon as a little salt is added, however, perfect agglutination occurs. To this technique the very obvious criticism may be applied that perhaps the absence of salt has precipitated the agglutinins, which, as we know, are precipitated with globulin, which is insoluble in the absence of salt. However, this source of error is excluded by the first experiment cited, and, moreover, it can be shown by the last experiment that, even though the bacteria are not agglutinated in the salt-free serum, they have nevertheless absorbed agglutinin. For, if such a salt-free mixture is centrifugalized, the bacteria washed and suspended in distilled water, and salt is then added, agglutination occurs. The supernatant fluid of the original suspension, further- more, can be shown to have been deprived of agglutinins by suitable experiment. These facts, first observed by Bordet, and further elaborated by the studies of Joos,62 Friedberger,63 and others, have been inter- preted in different ways. Joos claims that there is a chemical union between the bacteria and the salt, and bases this upon the observation that the salt added to a salt-free mixture cannot be demonstrated in 61 Bordet. Ann. de I'Inst. Pasteur, 1896, 1899. 62 Joos. Centralbl f. Bakt., 1, Vol. 33, 1903. 63 Friedberger. Berl. kl Woch., 1902; Centralbl f. Bakt., 1, Vol. 30. THE PHENOMENON OF AGGLUTINATION the supernatant fluid after agglutination has taken place. His ob- servations in this respect have not found confirmation at the hands of Friedberger and other workers, and it is generally agreed to-day that the role of the salts is, as Bordet first assumed it to be, a purely physical one. Bordet's opinion is often spoken of as the "two phase" theory, in that he conceives the process of agglutination to consist of two distinct occurrences, first, an absorption of the agglutinin by the bacteria, and, second, an agglutination of the new complex by the salt. It is not the agglutinin which causes agglutination, but by union with the agglutinogen forms a complex which is altered in "colloidal stability," and therefore is flocculable by the electrolyte. The opinion of Bordet becomes clearer as we consider the con- ditions governing the flocculation of colloids in general. Without wishing to enter in this place into detail regarding the nature of colloidal suspensions, it nevertheless seems necessary in order to do justice to this phase of the question to recall briefly the conditions governing such flocculation. The so-called colloidal solutions are not true solutions as the term is applied to dissociable substances, but are looked upon as consisting of small particles in suspension. The particles are similarly charged, as can be demonstrated by their wandering when subjected to an electric current, and it is supposed that it is this fact of similarity of charge which, in the asol" state, permits them to remain in suspension. For the similarity of the charges of the individual particles prevents their mutual approxima- tion. The state of suspension of these substances, then, represents a delicately balanced equilibrium between the two forces of electrical repulsion and of surface tension, an equilibrium which may be disturbed by the action of a number of factors. Thus, studies on inorganic colloids have shown, long before these considerations were applied to the explanation of serum reactions, that the stability of these suspensions could be disturbed both by electrolytes and by the addition of other colloids. Thus they may be precipitated by vari- ous salts, acids, and bases and, as Schultze 64 has shown, they react with that ion of the electrolyte which carries an opposite charge to that of the colloidal particles. For, although the colloidal units are similarly charged, this may be either negative or positive, according to the nature of the particular substance. In the case of the so-called amphoteric colloids reaction may take place, according to Pauli,65 with both ions of the electrolyte. The probable mechanism of the process is postulated by Pauli in describing the precipitation of a colloidal metal by salts, acids, or bases in the following way: 64 Schultze. Journ. f. prakt. Chem., 25, 1882, and 27, 1883. 65 Pauli. "Hofmeister's Beitrage," 1905, and "Physical Chemistry in Medicine," Wiley & Son, N. Y., 1907. INFECTION AND RESISTANCE "The introduction of such electrolytes into a colloidal suspension is of course accompanied by a certain amount of dissociation. In consequence the weakly charged particles of the colloid collect about the ions of opposite charge until a sufficient accumulation of the particles leads to an electrical neutralization of the ion, and the ag- gregation, if of sufficient size, will sink to the bottom, forming a precipitate." In regard to the mutual influences exerted upon each other when two colloids are mixed, it has been shown by Biltz, Hardy, and many other observers that oppositely charged colloids precipitate each other, though this is not an absolute rule, as experiments by Professor Stewart Young, of Stanford, have shown. Thus colloidal gold and platinum will be precipitated by such colloids as ferric oxid or aluminium oxid. When such a precipitating colloid is added to another oppositely charged suspension in quantities too small to bring about flocculation, moreover, the addition of a quan- tity of salt, likewise too small to precipitate alone, will in many cases bring about the flocculation. These and other phenomena of colloidal reaction have found close analogy in antibody-antigen studies, and have given support to the interpretation of the latter in the sense of Bordet. To return to the consideration of bacterial agglutination, we have spoken of the dependence of the reaction upon the presence of salts, and have seen that the researches of Friedberger and others have refuted the assumption that the action of the salt in bringing about agglutination depends upon chemical union of the salt with the bacteria. It is probable, therefore, that here, as in other colloidal precipitations, the function of the salt is to be regarded purely as an electrophysical phenomenon. The analogy becomes still closer when we consider the researches of Bechold,6'6 Neisser and Friedemann,67 Sears and Jameson,68 and others, which have shown that bacteria in suspension are to be com- pared very closely with true colloidal suspensions in that the bac- terial cells carry a definite and uniform electrical charge. Bacteria in salt solution emulsion, for instance, wander to the anode, thus giving evidence of their carrying a negative charge. This charge may be altered by adding to the emulsions definite con- centrations of acids or bases, a reversal of the charge taking place under the influence of NaOH or other hydroxids. Just how this is brought about is by no means clear, but it is not impossible that there is a selective absorption of OH ions by the bacteria, which therefore take on the charge of the ion. 66 Bechold. Zeitschr. f. physik. Chem., 48, 1904. 67 Neisser and Friedemann. Munch, med. Woch., Vol. 51, pp. 465, 827, 1904. 68 Sears and Jameson. "Thesis for M. A. Stanford University," 1912. THE PHENOMENON OF AGGLUTINATION 243 However this may be, and we must admit that explanations of these phenomena are as yet largely speculative, a fact which interests us particularly in connection with the phenomena under discussion at present is the influence exerted upon the charge of bacteria by exposure to the influence of serum. Bechold,69 as well as Neisser and Friedemann,70 assert that bacteria which have absorbed ag- glutinin no longer wander to the anode, but act as though they had been deprived of electrical charge, and precipitate between the elec- trodes. Bechold has suggested, for this reason, that it may be possible that bacteria in the normal condition are protected from the action of the electrolyte by a membrane or coating of protoplasm which acts as a protective colloid. The absorption of agglutinin may alter this in such a way that they become amenable to the flocculating effects of the salt ions. In keeping with such an opinion is the well- known observation of the inagglutinability of capsulated organisms, which, as Pprges 71 has shown, become agglutinable as soon as the capsules have been destroyed by heating in a weak acid. That the absorption of agglutinin indeed alters the electric sta- bility of the emulsified bacteria further appears from the fact that "agglutinin" bacteria 72 are agglutinated by concentrations of salts which are too slight to affect the normal micro-organisms. In this respect there is close similarity between the flocculation of agglutinin- bacteria and such emulsions as kaolin and mastic, whereas bacteria without agglutinin require much higher concentrations of the salts to produce like effects. The absorption of agglutinin may have re- moved a factor which protected the bacteria against the influence of the salt. On the other hand, it is equally just to assume — and this is more likely and corresponds with Bordet's views — that the ab- sorption of agglutinin has "sensitized" the bacteria to the action of the electrolyte. The experimental facts upon which the above state- ments are formulated are largely found in the important papers of Neisser and Friedemann — whose work brought out, likewise, inter- esting analogies of the colloidal precipitations with the phenomenon which we have described above as the proagglutinoid zone. In regard to the greater amenability of agglutinin bacteria to flocculation by electrolytes, the following protocol, adapted from the work of these authors, will explain itself. They were tabulated from experiments in which different quantities of normal y solution of various salts were added, on the one hand to emulsions of unal- 69 Bechold. "Die Kolloide in der Biologic u. Medizin," Steinkopf, Dres- den, 1912. 70 Neisser and Friedemann. Munch, med. Woch., Vol. 51, 1904, pp. 465 and 827. 71 Forges. Ztschr. f. exp. Path. u. Therapie, 1905. 72 "Agglutinin" bacteria — bacteria which have absorbed specific agglutinin. 244 INFECTION AND RESISTANCE tered bacteria, and, on the other, to bacteria which had absorbed agghitinin. It is seen that, with some salts, agglutination of the unaltered bacteria did not occur at all, whereas agglutination was brought about in the treated bacteria with comparatively small amounts; in other cases the- difference is a quantitative one only: Protocol constructed from the tables of Neisser and Friedemann, loc. cit. Y solution of salt Quantity of salt sol. which brought about agglutination of 1 c. c. of normal bacteria in emulsion. 0 = no agglutination by the salt solution Quantity of salt soL which agglutinated 1 c. c. of agglutinin bacteria in emulsion NaCl. . 0 .025 NaNO, o .025 Na2S04 0 .025 Rbl 0 . .025 MgS04 0 .0025 ZnS04 . . 01 001 CaCl2 0 .005 BaCl2 0 .005 Cd(N03)2 . . . 01 .001 CuS04 .0025 .0001 CuCl2 0025 .0005 Pb(N03)2 .0025 .0001 EfeCL.. .0025 .0005 The analogy between the experiment tabulated in the preceding protocol and the following from the work of the same writers is self- evident. Just as the absorption of agglutinin by bacteria rendered these more amenable to precipitation by salts, so the addition of minute quantities of gelatin to mastic emulsions had a similar sensi- tizing effect upon these. NaCl 10% solution 1 c. c. mastic (1-10 original emulsion) diluted to 3 c. c. 1 o. c. mastic + .0001 c. c. of a 2% gel. solution, the whole diluted to 3 c. c. 1.0 + + + + + + 0.5 0 + + + 0.25 0 + + + 0.125 0 + + + 0.05 0 0 0.025 0 0 Finally, one of the most important analogies yielded by the work of the above investigators is illustrated in the following protocol : THE PHENOMENON OF AGGLUTINATION 245 Colloidal iron hydroxid Precipitation emulsion of mastic 1 c. c. 1. 0 0.5 0 0.25 0 0.1 +- h 0.05 ++ + 0.025 ++ + 0.01 ++ + 0.005 ++ + 0.0025 +- h 0.001 0 Here we have an inhibition zone in the tubes containing the highest concentrations, accurately analogous to the previously dis- cussed proagglutinoid zone. It is a phenomenon similar also to the inhibition zones noticed in precipitin reactions and observed, though by a different technique, in bacteriolytic phenomena discussed in an- other place in connection with the Neisser-Wechsberg notion of com- plement-deviation or "Komplement Ablenkung." It seems to be a universal fact governing the union of colloidal substances, that defi- nite quantitative proportions must be maintained in order to lead to reaction, this being, possibly, explicable on the basis that actual union can take place only after disturbance of the electrical balance which keeps the particles apart. These reactions will be found more accurately discussed in another place. Whatever the mechanisms may be, however, these and similar experiments have seemed to render unnecessary and unlikely the assumption of proagglutinoids, proprecipitoids, etc., to explain the inhibition zones so frequently observed in all reactions of this kind. A peculiar observation, which does not coincide with the above interpretation of these phenomena, and the significance of which is indeed doubtful, is one which Friedberger 73 made in researches in which he confirmed the work of Bordet on the absence of agglutina- tion in a salt-free environment. He found that not only the addition of various salts would bring about agglutination under such condi- tions, but that organic substances such as dextrose and asparagin could be substituted for salts and had similar agglutinating effect— although higher concentrations of these than of the salts were re- quired. Were these substances at all dissociable it might be pos- sible that they acted by a mechanism identical with that of the salts — but since such substances as dextrose either do not dissociate at all or do so to an infinitesimal degree only there does not seem any pos- sibility of reconciling these results with Bordet's theory. 73 Friedberger. Centralbl. f. Bakt., 30, 1901. 246 INFECTION AND RESISTANCE It is difficult to explain Friedberger's results. Possible impurity of his preparations and the presence of traces of electrolyte seem to be excluded by the fact that he was quite conscious of this possi- bility of error and used only substances which yielded no ash on combustion. It may be that the results of Friedberger in which glucose and asparagin were used may have brought about agglutination by an entirely different mechanism from that which we are discussing and form no analogy to this. In one of the preceding paragraphs we have mentioned the phe- nomenon spoken of as "acid agglutination." By this is meant the spontaneous clumping, not only of bacteria, but of small particles of any kind, in suspension, in the presence of certain concentrations of acid. Michaelis,74 Beniasch,75 and others who have studied this phenomenon in detail have come to the conclusion that it is the con- centration of the hydrogen ions which is responsible for the ag- glutination. This explanation is also applicable to the agglutination often observed about the anode when bacteria are subjected in sus- pension to the action of a direct current. In such experiments the organisms after concentrating at this electrode often flocculate, and it is here, of course, that hydrogen ions are present in the greatest concentration. How this takes place is problematical, but the reason- ing of Pauli, if applied to this, would favor the assumption that the weakly charged bacteria group themselves about the ions and, when a sufficiently large aggregation has formed, fall to the bottom as precipitate. This phenomenon of acid agglutination is of course entirely different in nature from the specific serum agglutination which we are discussing. Nevertheless, Schidorsky and Reim,76 Jaffe,77 and others have attempted to apply acid agglutination to the isolation and differentiation of bacteria, on the conception that dif- ferent species are agglutinated by varying concentrations of hy- drogen ions. The former investigators, even, claim to have been successful in isolating typhoid bacilli from the stools by this method in that the typhoid bacillus was agglutinated by concentrations of acid which had no effect upon the Bacillus coll. Sears 78 has gone over this work carefully, and, while he has obtained results which bear out the contention that the agglutination is probably due to the concentration of the H ions, his experiments have revealed an irregu- larity in the behavior of bacteria of the same species in acid solutions and an overlapping of those of one species with those of another. Therefore the use of acid agglutination for differential purposes 74 Michaelis. Folia Serol, 7, p. 1010, and Deutsche med. Woch., 37, 969. 75 Beniasch. Zeitsckr. f. Imm., Vol. 12, 1912. 76 Schidorsky and Reim. Deutsche med. Woch., Vol. 38, p. 1125. T7 Jaffe. Arch. f. Hyg., Vol. 76. 78 Sears. Proc. Soc. of Exp. Biol. and Med,, 1913. THE PHENOMENON OF AGGLUTINATION seems to us entirely hopeless. And indeed it would be surprising if any such distinctive and regular reaction differences between simple bacterial cells, after all chemically and physically so essentially alike, could be found. CHAPTER X THE PHENOMENON OF PRECIPITATION (Precipitins) THE establishment of the agglutinin reaction as a constant and specific serum-phenomenon by the work of Gruber and Durham led immediately to assiduous investigation of the many problems sug- gested by it, and among them, as we have seen, the question of the nature of the agglutinogen. It was found that agglutinins could be produced, not only by the injection of whole bacteria, but equally as well by treatment with dissolved bacterial extracts or with filtrates from old broth cultures. This naturally led to the thought that there might be a definite reaction if such extracts (instead of the bacteria themselves) were added to agglutinating sera in vitro. Rudolf Kraus l was the first to perform this very logical experiment. He was working with broth filtrates of Bacillus pestis and of the cholera spirillum, and found that when he mixed the perfectly clear filtrates of such cultures with their respective antisera the mixtures would at first become turbid and finally show a light flocculent precipitate. He named the reaction the "precipitin reaction" and, in analogy to agglutinins, spoke of the bodies in the serum which caused the pre- cipitation as "precipitins." The reaction was found, like that of agglutination, to be specific; the cholera serum gave no precipitate with the plague extract and vice versa, and Kraus, after extending his observations to other bacteria, pointed out the practical diagnostic possibilities of his discovery. Though Kraus' first observations were made entirely with bac- terial culture filtrates and antibacterial sera, it was soon discovered that his results were merely isolated instances of a broad biological law, and that specific precipitins were produced whenever animals were treated with injections of any kind of foreign protein. Thus Tschistovitch,2 in 1899, found that the blood serum of rabbits im- munized with eel-serum gave specific precipitates when mixed with eel-serum, and Bordet 3 obtained analogous results by treating rab- bits with defibrinated chicken blood and with milk. Thus rapidly 1 R. Kraus. Wien. Idin. Woch., No. 32, 1897. 2 Tschistovitch. Ann. de I'Inst. Past., 13, 1899. 3 Bordet. Ann. de I'Inst. Past., Vol. 13, 1899, pp. 225-273. 248 THE PHENOMENON OF PRECIPITATION 249 the discovery of Kraus was developed into the generalization that the sera of animals that have been treated with foreign proteins of any kind — bacterial, animal, or vegetable — will develop the property of causing precipitates when mixed with clear solutions of the re- spective antigens. The substances which, after injection into the animal body, lead to the formation of precipitating antibodies are spoken of in the language of immunology as "precipitinogen." In the case of bac- teria it has been shown that, while the injection of the whole bac- terial cell — dead or alive — will lead to precipitin formation, bacterial extracts produced in a variety of ways will lead to the same result. Such precipitinogen extracts can be obtained by allowing the bacteria to grow in flasks of slightly alkaline bouillon, keeping them in the incubator for from three weeks to three months, and then filtering them through Berkefeldt candles. Again, useful extracts can be more rapidly produced by growing large quantities of bacilli on agar, emulsifying in salt solution, and shaking in any one of the ordinary types of shaking machine for 48 hours or longer. On filter- ing an extract is obtained which will form precipitates with homol- ogous immune serum, or will incite precipitins when injected into animals. In fact, any one of the customary vigorous methods of extracting bacterial or other cells will yield precipitinogen. A rela- tively purified precipitinogen in the form of a dry, water-soluble powder has been obtained by Pick by the precipitation of culture filtrates with alcohol. Regarding the chemical nature of the precipitin-inducing sub- stances, or precipitinogens, the same problems have arisen which have been discussed in connection with antigens in general. We may say that all soluble native proteins possess precipitin-inducing prop- erties. Yet this does not sufficiently define the term, since many observations have been published which show that physically and chemically altered proteins may still induce specific precipitins; a few investigators, furthermore, have claimed that they have produced non-protein precipitinogen by various methods of breaking up the molecule of the original antigen. In the section on agglutination we have seen that moderate heating (56-65° C.) rather increases than decreases the agglutinogen characteristics of bacteria, and it is equally true that such heated bacteria or bacterial extracts may induce precipitins. However, regarding the action of higher de- grees of heat (boiling) upon precipitinogens in general we will have more to say in another place. Of more immediate, indeed of fundamental, importance is the problem of a non-protein antigen. The most important claims in this regard have been made by Pick,4 Obermeyer and Pick,5 and by 4 Pick. "Hofmeister's Beitrage," Vol. 1, 1901. 5 Obermeyer and Pick. Wien. klin. Woch., 1904, p. 265. V 250 INFECTION AND RESISTANCE Jacoby. 6 7 Jacoby, working with a vegetable antigen, ricin, found that by trypsin digestion he could obtain a substance which still retained antigenic properties, but no longer gave any of the pro- tein reactions. Obermeyer and Pick, by the same method, claim that they have produced a non-protein precipitinogen from egg al- bumen. On the other hand, others have had negative results, and Kraus8 himself, after reviewing the evidence on both sides, comes to the conclusion that available data do not justify us in separating the antigenic properties from the protein molecule. In unpublished experiments which the writer carried on in the laboratory of Profes- sor Friedemann in Berlin also attempts to produce a non-protein precipitinogen from horse serum by bacterial putrefaction were en- tirely negative. The putrefaction of the serum, though carried out in dialyzing bags for the removal of diffusible products, was ex- tremely slow, and when finally the Biuret reaction disappeared the serum was no longer precipitable by potent antisera. However, the flaw in these experiments is that the true test of the presence of precipitinogen is not the precipitable character of the solution in question, since actual precipitation is dependent, as we shall see, upon many modifying secondary factors, but rather the ability of the substance to induce precipitins in treated animals. The fact that Mcolle,9 and later Pick,10 were unable to obtain alcohol-soluble substances from bacteria and bacterial extracts which were still precipitable might also be taken to point toward the non- protein character of the precipitinogens, suggesting that these sub- stances may be of a lipoidal nature. However, as Landsteiner n points out, mere solubility in organic solvents can no longer be taken as a proof of lipoidal character, since it is more than probable that non-lipoidal substances may go into alcoholic and other organic solu- tion when lipoids, such as lecithin, are present. Thus Miiller 12 found that the antigen of typhoid bacilli was soluble in chloroform in the presence of old preparations of lecithin. Pick and Schwartz,1 3 who had previously studied similar antigen solubilities in the pres- ence both of lecithin and of other organ lipoids, suggest that possibly such solutions represent lipoid-protein combinations — colloidal "so- lutions"— which permit the presence of protein mechanically or chemically united to the lipoid in the organic solvents — alcohol, chloroform, etc. Here, too, then there is no evidence for the ex- istence of non-protein precipitinogen. 6 Jacoby. "Hofmeister's Beitrage," Vol. 1, 1901. 7 Oppenheimer. "Hofmeister's Beitrage," Vol. 4, 1904, p. 259. 8 Kraus in "Kolle u. Wassermann Handbuch," Vol. 4, p. 605. 9 Nicolle. Ann. de I'Inst. Past., 12, 1398. 10 Pick. "Hofmeister's Beitrage," Vol. 1, 1901. 11 Landsteiner. "Weichhardt's Jahresbericht," Vol. 6, 1910, p. 214. 12 Miiller. Zeitschr. f. 1mm., Vol. 5, 1910. 13 Pick and Schwartz. Biochem. Zeitschr., Vol. 15, 1909. THE PHENOMENON OF PRECIPITATION 251 Of importance in connection with the problem of the nature of precipitinogen, also, is the claim of Myers/4 that specific precipitins may be produced in rabbits by treatment with Witte peptone, a sub- stance complex in constitution, but consisting largely of albumoses. This observation has failed of confirmation in the hands of Ober- meyer and Pick, Michaelis,15 Norris,16 and others, and cannot, there- fore, be accepted as an established fact. Whichever method of precipitinogen production is used bacterial precipitins appear in the serum of the immunized animal only after careful and continued immunization, usually later than the demon- strable appearance of the bactericidal or agglutinating properties of the serum. The most convenient material for such immunization^ consists of salt solution emulsions of agar cultures, killed at 60° to 70° C. These may be injected subcutaneously, intraperitoneally, or intravenously, the last method leading to the most satisfactory and rapid results and, therefore, best employed unless great inherent toxicity of the particular bacteria contraindicates. When rabbits are used it is generally necessary to inject 3, 4, or 5 times at 5 or 6- day intervals, and to bleed the animals on the 8th or 9th day after the last injection. The bacterial precipitins so produced are, as we have said above, specific — but, again, specificity, as in the case of agglutinins, is limited by the so-called "group reactions." In the chapter dealing with agglutination we have seen that the serum of a typhoid-immune animal which agglutinates typhoid bacilli strongly will also aggluti- nate, though far less powerfully, paratyphoid bacilli and, in some cases, even colon bacilli, this appearance of "minor" agglutinins being probably due to a close group relationship of these bacteria to the typhoid bacillus. In the case of bacterial precipitins the same thing is true, and has been made the subject of special studies by Zupnik,17 Kraus,18 Norris,19 and others. As in the case of ag- glutination, however, this fact does not in any way interfere with the practical value of the specificity of the reaction because elimina- tion of the secondary group reactions, which in agglutination is obtained by dilution of the antiserum, can here be obtained, as Kraus points out, by diminishing the quantity of the undiluted precipitat- ing serum added to the bacterial filtrates. Thus, while one volume of serum added to one, two, or three volumes of culture filtrate may still give error due to non-specific group reactions, a proportion of "Myers. Centralbl f. Bakt., Vol. 28, 1900. 15 Michaelis. Deutsche med. Woch.j 1902. 16 Norris. Jour, of Inf. Dis., Vol. 1, 1904. 17 Zupnik. Zeitschr. f. Hyg., 49, 1905. 18 Kraus. Wien. klin. Woch., 1901, No. 29. 19 Norris. Jour, of Inf. Dis., Vol. 1, 1904. 252 INFECTION AND RESISTANCE one part of serum to 8 or 10 parts of the filtrate will usually elimi- nate all secondary reactions and prove strictly specific. An illustration of such an elimination of "partial" or "minor" precipitins by diminution of the amount of the homologous anti- serum is given in the following table taken from the work of Kor- ANTICOLI RABBIT SERUM TABLE III The precipitating action of the anticoli rabbit serum upon its corresponding filtrates and upon the filtrates of B. N° 1 (hog cholera) and B. typhosus. Coli filtrate Anticoli aerum 0.5 c. c. 0.05 Cloudiness in all tubes in 1 hour at 37.5° C. which 0 . 5 c. c. 0 . 10 increases rapidly. Six hours well-marked precipita- 0.5 c. c. 0. 15 tion — most copious in tube containing 0.25 serum. 0.5 c. c. 0.25 Fluid in all tubes becomes clear. B. N°l filtrate Anticoli serum 0 . 5 c. c. 0 . 10 At 6 hours a slight precipitate in the form of fine 0.5 c. c. 0.25 granules appears on the sides of the tubes. After 24 hours the precipitate in the tube containing 0.25 c. c. serum compares in amount to that formed in the homologous filtrate with 0.05 c. c. of serum. B. typh. (Coll) filtrate Anticoli serum 0.5c. c. 0.10 Similar reaction obtained to that with B. N° 1 filtrate. 0.5 c. c. 0.25 B. typh. (Pfeif- fer) filtrate 0.5c. c. 0.10 Similar delay in reaction as obtained with B. typh. 0.5 c. c. 0.25 Coll. And, indeed, though the great practical value of the precipitin reaction has not been in the special field of bacteriology, it has been successfully utilized in the diagnosis of glanders by Wladimiroff,21 and constitutes a valuable adjuvant to our methods of determining the biological relationship between micro-organisms. The production of precipitins against unformed proteins, egg albumen, animal sera, etc., is much more easily accomplished than the production of bacterial precipitins, and three intravenous injec- tions of from 2 to 5 c. c. of the protein at 5 or 6-day intervals usually give rise to a formation of potent precipitins. When a small quan- tity of the serum of such an animal, taken 9 or 10 days after the third injection, is mixed in a test tube with an equal quantity of a 20 Norris. Jour, of Inf. Dis., Vol. 1, 1904, p. 472. 21 Wladimiroff. "Kolle u. Wassermann Handbuch," article on "Glanders/' Vol. 5, 2d Ed. w PHENOMENON OF PRECIPITATION 253 dilution of the protein which was injected, turbidity and rapid floccu- lation will result. In tests of this kind, unlike the bacterial precipitin tests in which the delicacy of the reaction is ordinarily determined by diminution of the amounts of antiserum, the same object may be more conveniently attained by dilution of the antigen. Thus, in test- ing the precipitating potency of, let us say, the serum of a rabbit immunized with sheep serum, we would proceed by setting up a series of small tubes, each of which contains a constant amount of antiserum (precipitin), but a progressively diminishing amount of antigen in the same volume — i. e., in dilution with isotonic salt solu- tion. The following example will make this clear : Anti sheep serum from rabbit Sheep serum 0.5 c. c. of following dilutions: Precipitation 0.5 c. c. + •10 ± 0.5c. c. + :100 + + + 0.5 c. c. + :500 + + + 0.5 c. c. -f :1,000 + + 0.5 c. c. -f :5,000 -|- 0.5 c. c. + :10,000 — In this test it will be noticed that the strongest concentration of the antigen (1:10) gave a relatively slight precipitation only. This phenomenon is analogous to the inhibition zones noticed in the case of agglutination and other antibody reactions and will be more espe- cially discussed in a succeeding paragraph. The delicacy of the above example, moreover, is by no means unusual, and sera have been obtained by careful immunization with which the specific antigen could be detected in dilutions as high as 1 to 100,000 (Uhlenhuth). A serum which will react with antigen dilutions of 1 to 10,000 and 1 to 20,000 is not at all uncommon nor difficult to obtain. Apart from the additional advantage of the specificity of the reaction, therefore, this biological method of de- tecting proteins is more delicate than that of any of the known chemical methods; neither the Biuret nor Millon's reaction will far exceed a delicacy of 1 to 1,000. By a modification known as the method of Complement or Alexin-fixation, furthermore, the delicacy of the biological reactions can be still further enhanced. This is discussed in detail in another place (see page 212). The practical value of the precipitin reaction, however, lies, not in the mere detection of protein, but, by virtue of its specificity,22 in the determination of the variety of protein under examination. And 22 Wassermann and Schiitze. Deutsche med. Woch., 1900, Vereinsbeilage, p. 178; Berl. kl Woch., 1901; Deutsche med. Woch., 1902; Bordet, Ann. Past., Vol. 13, 1899; Nolf, ibid., Vol. 14, 1900; Fish, Medical Courier, St. Louis, 1900, cited from Wassermann. 254 INFECTION AND RESISTANCE here again the specificity, like that of bacterial precipitation, ag- glutination, and other serum tests, is relative rather than absolute. Thus a serum which has been obtained by the immunization of an animal with human serum may react, not only with human serum, but also with relatively higher concentrations of the sera of some of the higher apes. However, such non-specific partial reactions can be eliminated entirely by employing higher dilutions of antigen. Thus Uhlenhuth,23, 24, 25 on the basis of a large experience, has established a standard of antigen dilution at 1 to 1,000, beyond which no "para" or "minor" precipitation will occur. Since potency far exceeding this is easily procured, absolute specificity can be ensured by the very simple precaution of a sufficient dilution. The most important practical use for the reaction has been found in forensic medicine, where it is possible in this way to determine the species of animal from which have emanated the blood, sperm, etc., found in spots on wearing apparel, weapons, or other articles. The extensive investigations of Nuttall 26 upon this subject have inci- dentally been of much value in furnishing a further method for the determination of zoological species relationships. Nuttall carried out 16,000 precipitin tests, with precipitating sera, upon 900 speci- mens of blood which he obtained from various sources. He not only confirmed many of the accepted zoological classifications, but shed much light upon a number of disputed points. In working out the tests upon monkeys he found that the reactions carried out with anti- human serum become weaker as the species examined is farther re- moved from man zoologically. Thus as we read down the column from man to the hapalidaB the precipitate becomes less and less in amount. Nuttall's Tests with Antihuman Serum. (Nuttall, loc. cit., p. 165.) ANTIHUMAN PRECIPITATING SERUM Tested against Precipitate 34 Specimens human blood 100%27 8 Simiidse, 3 species 100% 36 Cercopithecidse 92% 13 Cebida? 78% 4 Hapalidae 50% 2 Lemuridse 0 23 Uhlenhuth. Deutsche med. Woch., 1900, 1901 ; Bob. Koch Festschrift,. 1903. 24 Uhlenhuth and Weidanz. "Kraus u. Levaditi Handbuch," etc., Vol. 2, 1909. 25 Uhlenhuth and Weidanz. Loc. cit., where other publications are sum- marized. 26 Nuttall. "Blood Immunity and Blood Relationship," Cambridge Uni- versity Press, 1904. 27 The percentages refer to the volume of precipitate formed on standing for a given time, the amount formed by the antiserum with its specific antigen being taken as 100 per cent. Antigen dilutions correspond throughout. THE PHENOMENON OF PRECIPITATION 255 In another series he finds: ANTIHUMAN PRECIPITATING SERUM Tested against Precipitate Man 100% Chimpanzee (loose precip.) 130% Gorilla 64% Ourang 42% Cynocephalus mormon 42% Cynocephalus sphinx 29% Ateles 29% Among the primates the highest figures with antihuman serum are given by the chimpanzee. Other bloods than those of the primates • gave slight reactions or none whatever with the antihuman serum. In addition to these results the relationships within the dog family, the horse family, and many other kinships similar to these were confirmed. In every case the precipitin reaction was con- sistent with the results of other methods of classification, and ^N"ut- tall's work is an extremely valuable aid to zoologists in disputed questions of animal relationships. These facts are the more surprising in that they demonstrate species differences between the proteins of various animals which are not determinable by known chemical methods. How funda- mental these differences are and how delicate the reaction, is further shown by experiments of Uhlenhuth, in which he obtained a specific antihare serum by treating rabbits' with hares' blood, an astonishing result in view of the close zoological relations between these animals. Isoprecipitins, that is, precipitins resulting from the treatment of animals with blood of another individual of the same species, have also been described by Schiitze and others. They are not, however, regular in their appearance, nor are they very potent when obtained. Since the reaction is equally applicable to vegetable proteins, similar investigations on the interrelationship of different varieties of wheat have been carried out by Magnus.28 The methods of performing precipitin tests for forensic or other purposes is extremely simple. Nevertheless, there are a number of theoretical considerations which we must take up in order to make clear the limitations of accuracy and conditions of control which are involved in these reactions. From our discussion of the nature of precipitinogen it follows that blood stains, etc., on linen or articles of any kind will be suitable for precipitin tests even after they have been exposed for considerable periods to unfavorable conditions, that is, an environment in which they are subjected to exposure to light, moderate heat, or drying. Thus blood spots, etc., if kept dry and in 28 Magnus. Cited from Uhlenhuth, loc. cit. 256 INFECTION AND RESISTANCE the dark, may give positive reactions even after years, as experi- ments by Uhlenhuth have shown. Meyer 29 claims even to have obtained a precipitation with extracts of the material of mummies. One of his specimens was a mummy dating back to the first Egyptian Empire (5,000 years), the other about 2,000 years old. Pieces of the leg and neck muscles of these specimens were chopped up finely, extracted for 24 hours with salt solution, then filtered until clear. With antihuman serum they gave turbidity after one hour at 37.5° C. Under conditions of putrefaction, of course, the precipitinogen is more rapidly destroyed, though blood putrefies with surprising slowness, even if, as in our own experiments, the conditions of mois- ture, temperature, and reinoculation with putrefactive bacteria are constantly observed. Under such conditions a weak reaction may be obtained after as long as a month or six weeks. In carrying out the tests with any material it is first necessary to get it into clear solution, a result which is best accomplished by soaking it in a small quantity of isotonic salt solution. Preliminary to this it. is always necessary to scrape off a bit of the specimen and examine it microscopically to discover, if possible, whether blood cells, sperm, or other cellular constituents can be detected. The infusion in salt solution should be continued for several hours — if necessary for 12 to 24 hours. After the first few hours in the incubator the material should be placed at room or refrigerator temperature so that the yield in unchanged protein may not be diminished by the action of bacterial growth. After extraction the solution may be filtered in order to clear it, but often mere centrifugation suffices for this pur- pose. The concentration of antigen in such an extract is always an uncertainty, but may be determined with sufficient accuracy for practical purposes by shaking and observing the formation of a lasting foam. Protein solutions will show foam on shaking in dilu- tions as high as 1 to 1,000, and if the original amount of salt solution used in washing out the material is properly gauged to the amount of blood available in the stain, and the solution shaken and observed for the formation of foam, it is usually a simple matter to obtain a final concentration approximating one to one thousand.30 The antiserum which is used should be of such a potency that preliminary titration with the specific antigen, diluted 1 to 1,000, should give an almost immediate cloudiness at room temperature. By testing this serum against a number of other varieties of 29 Meyer. Munch, med. WocK, Vol. 51, No. 15, 1904. 30 If there is enough material, the amount of dissolved protein can be also approximately gauged by adding to a little of it a drop of acid, boiling and observing the heaviness of the cloud which forms. A control test of a known dilution of the suspected variety of blood can be made at the same time and the heaviness of this cloud compared with that in the test solution. THE PHENOMENON OF PRECIPITATION 257 protein — dog serum, beef serum, etc. — it must be determined that the precipitin in this case is strictly specific. The reaction can be observed with greater delicacy if it is first set up by the method recommended by Fornet and Miiller,31 which we may speak of as the "ring test." The antiserum is put into the tubes and the solution to be tested is allowed to flow slowly over this —as in Heller's nitric acid albumin test. At the line of contact be- tween the two a fine white ring will rapidly appear, thickening and growing heavier as the preparation is allowed to stand. After taking the final readings from such a test, let us say after an hour or so, it is well to shake up the tubes, set them away in the ice-chest, and again read the amount of precipitates formed in the various tubes the next morning. Since every test of this kind necessitates a number of controls, the following example will serve as a basis for discussion : Forensic Blood Examination Material: Blood spot on trouser pocket, washed up in salt solution. Clear after paper filtration. Antiserum: Rabbit treated with three intravenous injections, 2, 5, and 5 c. c. of human serum at six-day intervals; bled on tenth day after last injection. This serum has been titrated against human serum and gives precipitation in dilutions up to one to ten thousand. With one to one thousand there is clouding which begins in three minutes and is very distinct in eight minutes, at room temperature.32 Test Tube 1. Known human serum 1 to 1,000. . 1.0 c. c. + Antiserum. . . .0.2 c. c. Tube 2. Unknown solution to be tested 1.0 c. c. -f Antiserum. . V0.2 c. c. Tube 3. Unknown solution to be tested 1.0 c. c. + Normal rabbit serum 0.2 c. c. Tube 4. Salt solution 1.0 c. c. + Antiserum .... 0.2 c. c. Tube 5. Unknown solution 1.0 c. c. -f Salt solution. . .0.2 c. c. In this test, if the original material was human blood, tubes 1 and 2 should show ring formation within 5 minutes — while the other tubes remain clear. In addition to these controls it is well to be sure that the test extract is neither strongly acid nor alkaline, and that, as Uhlenhuth suggests, the material from which it is extracted does not contain other substances which can give precipitates by themselves when added to serum. This is especially necessary in the case of cloth fabrics, and a recent instance in our own experience has suggested to us the possibility that such materials may also con- tain colloidal dye stuffs or other extractable substances which can cause inhibition of the precipitation. In an apparently positive case the reactions with a blood extract from trouser cloth were suffi- ciently heavy, but regularly delayed, as in the flocculation of such 31 Fornet and Miiller. Zeitschr. f. Hyg., Vol. 66, 1910. 32 A mixture of too specific antisera should never be used, since such sera may often precipitate each other for reasons that are discussed below. 258 INFECTION AND RESISTANCE colloidal suspensions as arsenic trisulphide in the presence of a protective colloid. In the ordinary criminal or civil case which would come under consideration for precipitin tests the spots or stains are made by blood as it flows from the wound and unchanged by chemical or physical agencies except as these are encountered afterward, by exposure. In the case of meat inspection, in which the precipitin test is useful in detecting admixtures of horse flesh, dog flesh, or other less desirable varieties of meat, in sausages, chopped meat, etc., it often happens that such procedures as heating or smoking may vitiate the results of precipitin reactions. It is of practical im- portance, therefore, that we should know exactly what the effects of heating (boiling) may be upon precipitinogen. Moreover, this ques- tion possesses considerable theoretical interest since the coagulation of proteins by heat seems to involve chiefly a physical rather than a chemical change. Cohnheim 33 says in discussing this question : "It is still unclear what the changes are that take place in coagulation. It may be that there is merely an intramolecular 'Umlagerung' — or there may be cleavage ; or the process may be comparable to the flocculation of col- loidal clay emulsions by salts. . . . With coagulation all proteins have lost the differences which they possess in the native state in respect to solubility or precipitability by salts. Physically all coagu- lated proteins are alike ; they are no longer native proteins, and with- out further decomposition are insoluble. Chemical differences, how- ever, variations of composition, and the cleavage products which they yield still distinguish them." The question has been experimentally approached by Obermeyer and Pick 34 in connection with their general investigations upon the influence of chemical and physical alterations upon precipitinogen. They found that precipitin produced with unchanged (native) beef serum does not react with heated beef serum, even if immunization was prolonged and a very potent serum was produced. On the other hand, when animals were immunized with beef serum which had been boiled for a short time ("Kurz auf gekocht" 35 ) the precipitin so produced reacted, not only with native beef serum, but also pre- cipitated the boiled serum and a whole row of split products which give no reaction to normal precipitin. The "coctoprecipitin" so produced, furthermore, was found by them to be specific, acting only upon beef protein or its derivatives. 33 Otto Cohnheim. "Chemie der Eiweiss Korper Vieweg Braunschweig," 1900, p. 8. 34 Obermeyer and Pick. Wien. kl. Woch., 12, 1906. 35 Sera or other proteins used in such tests are boiled in dilutions of 1 to 10 or more, in order to avoid the formation of heavy flakes which cannot be injected. Boiled in sufficient dilution, an opalescent suspension is formed which easily passes through the syringe. THE PHENOMENON OF PRECIPITATION 259 It is immediately evident that these investigations are closely analogous to those of Joos and others on the agglutinins. The anti- serum produced with the heated antigen here again reacts both with the native and with the heated antigen, whereas the antiserum pro- duced with the native unheated antigen reacts only with the un- heated. The "heat-precipitins" therefore may be also called "um- fanglicher" — the term applied by Paltauf to the agglutinins pro- duced with heated bacteria. Schmidt,36 who has studied the problem extensively, finds that heating serum protein to 70° C. for as long as 30 to 60 minutes alters its precipitability by "native precipitin" (precipitin produced by immunization with native unheated serum) only in so far as it diminishes the delicacy of the reaction by 10 to 30 per cent., and that heating to 90° C. for as long as an hour does not render it en- tirely non-precipitable, so that protein so treated may yet be detect- able by ordinary specific precipitins produced by injections of un- heated serum, though the delicacy of the reaction is lessened. Boil- ing, according to Schmidt, renders the antigen no longer precipitable by such "native precipitin," but, on the other hand, it does not seem to destroy its antigenic property of inciting precipitins on injection into animals. Fornet and Miiller, on the other hand, claim that even boiled protein can be detected by "native precipitins," though the reaction is only about one-tenth as delicate as it is with unheated protein. Schmidt studied these relations especially as they affect the per- formance of specific precipitin reactions in the identification of boiled meat. He found that when he immunized rabbits with serum protein that had been heated at 70° C. for 30 minutes the antiserum so obtained gave strong and practically useful reactions with its specific antigen even if this had been boiled. Since "native pre- cipitin" gives weak reactions only with such a boiled protein, Schmidt recommends the use of the "TO0 precipitin" (produced by injections of heated serum) for tests in which a heated antigen is to be identified. He states, however, that very prolonged heating may so com- pletely coagulate the antigen that none of it can be gotten into "solu- tion" (suspension), and in such cases results can be obtained neither with the "native" nor with the "70° precipitin." He has at- tempted, therefore, to find a method whereby even such entirely insoluble proteins may be identified, and claims to have succeeded by preparing what he calls his "heat-alkali-precipitin." 37 He di- 36 Schmidt. Biochem. ZeitscJir., 14, 1908; also Zeitschr. f. Imm., Vol. 13, 1912. 37 "Native precipitin" = precipitin produced by injections of normal un- heated serum. "70° precipitin" = precipitiif produced by injections of serum heated to 70° C. for 30 minutes. 260 INFECTION AND RESISTANCE lutes serum with equal parts of isotonic salt solution and heats it to 70° C. for 30 minutes in a water bath. To 60 c. c. of such a solution he now adds 10 c. c. of ? NaOH, and continues heating for 15 to 20 minutes. At the end of this time he neutralizes with HC1, cools, and injects 20 c. c. intraperitoneally into rabbits. (The neutraliza- tion is not absolutely necessary.) Five or more injections yield a serum sufficiently potent for use. A precipitin so produced will, according to Schmidt, react spe- cifically with heated proteins, and also with protein which has been solidly coagulated and brought into solution by means of N"aOH and heat. It will not, however, react with normal unheated antigen. He tested this by coagulating horse serum by boiling for 3 hours. The coagulum was washed with salt solution, dried, and powdered. Tests were then made to prove that this powder was entirely in- soluble in !N~aCl solution. A little of it was then treated with 10 c. c. of salt solution containing enough NaOH to correspond to an ^ solution. The exposure was continued for 20 minutes in a water bath at 60° to 70° C. Before the entire mass was dissolved the solu- tion was filtered and neutralized with -^ HC1. The rather complicated relations described by Schmidt are easily surveyed in the following protocol taken from his work : TABLE I (W. A. Schmidt, Zeitschr. f. 7mm., Vol. 13, 1912, p. 173) Solution of Native precipitin Heat (70°) precipitin Heat-alkali- precipitin Native serum Strong reaction Good reaction 0 (very slight 70° serum (heated 30 min.) 100° serum (heated 30 min.) Good reaction 0 Strong reaction Good reaction turbidity) Strong reaction Strong reaction 70° serum treated with NaOH (used to produce heat-alkali-precipitin) . Boiled insoluble serum, brought into solution with NaOH 0 0 0 o Strong reaction Good reaction Native serum treated with NaOH in the cold 0 0 Good reaction Schmidt speaks of the "heat-alkali-precipitin" also as "alkali- albuminate-precipitin." It can be produced only if the NaOH treat- ment of the serum is cautiously performed. If the sodium hydroxid is allowed to act too vigorously in strong concentrations or for too long a time the antigen is completely destroyed, is no longer pre- THE PHENOMENON OF PRECIPITATION 261 cipitable, and no longer produces precipitin when injected into animals. The striking feature of these experiments is that they show a gradual alteration of the protein first by heat, then by alkali and heat, in such a way that the antigenic properties are changed but not destroyed. Each precipitin, moreover, seems to react most strongly with the particular antigen-alteration which produced it, and, according to Schmidt, retains its species specificity. This is not the case with the iodized proteins and nitroproteins and diazo- proteins produced by Obermeyer and Pick.38 Here iodized beef protein injected into animals produced a precipitin which reacted with the iodized protein, not only of the beef, but also similarly altered proteins of other animals — and the same was true of the nitro and diazo modifications. Although the experiments of Schmidt have great theoretical value, their practical utilization must depend upon the degree of specificity possessed by the heat-precipitins or the heat-alkali-pre- cipitins. In Obermeyer and Pick's original investigations we have seen that they found the precipitin produced with heated serum as strictly specific as that induced by native serum. This has also been the experience of Schmidt. Fornet and Miiller,39 on the other hand, report that the precipitins produced by them with heated muscle- protein were not as strictly specific as those produced with the un- heated — in that the former gave precipitates, not only with homol- ogous protein solutions, but with foreign proteins in moderate con- centration as well. In experiments carried out by the writer with Ostenberg 40 it was attempted to determine whether or not precipi- tins could be produced by injecting animals with protein that had been boiled, and if so what the action of these substances would be upon boiled proteins. Contrary to the results of Fornet and Miiller, it was actually found that sera boiled for 3 to 5 minutes injected into rabbits induced precipitins which acted upon boiled proteins, but at the same time it was determined that the antibodies so produced were no longer strictly specific. The protocol given at the top of the next page will illustrate these experiments. Summarizing these results together with those of Fornet and Miiller and of Schmidt it would seem that the injection of boiled proteins induces precipitins which no longer act on native antigen, which act powerfully on boiled antigen, but are no longer strictly specific. This seems to us of great theoretical interest as showing an alteration by heating in the species adherence of the antigen. Practically, therefore, precipitins produced with boiled protein are of little value, and forensic determinations of boiled proteins should 38 Obermeyer and Pick. Wien. klin. Woch., No. 12, 1906. 39 Fornet and Miiller. Zeitschr. f. Hyg., Vol. 66, 1910. 40 Zinsser and Ostenberg. Proc. N. Y. Pathol Soc., 1914. 262 INFECTION AND RESISTANCE Experiments on Cocto-precipitin. Table II (March 23, 1913). Cross titrations — dilutions of sera in salt solution boiled 5 minutes, precipitated with antisera produced by injections with similarly boiled material. The readings here indicated were taken by "ring" test at the end of 30 minutes. Beef Beef Beef Dog Dog Dog Sheep Sheep Sheep serum serum serum serum serum serum serum serum serum vs. vs. vs. vs. vs. vs. vs. vs. vs. Dilution anti- anti- anti- anti- anti- anti- anti- anti- anti- beef precipi- dog precipi- sheep precipi- dog precipi- beef precipi- sheep precipi- sheep precipi- dog precipi- beef precipi- tin tin tin tin tin tin tin tin tin 1:20 + + + + + + + + + + + 1:50 + + + + + + + + + + + + + + + + + + 1:100 + + + + + + + + + -f + + + 1:500 + + + + + + 1:1,000 ± — ± ± ± Controls of boiled serum alone* 1:20 1:50 Serum control * These controls were necessitated by the fact that the boiled serum suspensions were them- selves turbid and occasionally showed slight settling on standing. be done, as advised by Schmidt, by the "70° precipitins," or with native precipitin as practiced by Fornet and Miiller. The specificity which is the basis of the practical value of the reac- tions that we have discussed is spoken of as "species" specificity since it iias been found that the blood serum of rabbits or other ani- mals into which the serum of another animal has been inji^ed reacts, not only with the homologous blood serum, but also with extracts of the various organs of the particular species of animal which furnished the serum. Thus if we immunize rabbit, let us say, with sheep serum the resulting precipitin will react, not only with sheep serum, but also with extracts of sheep spleen, sheep liver, etc. It seems that every species of animal possesses throughout its tis- sues a particular variety of protein, fundamental to its general metabolism and peculiar to its species. On the other hand, we have seen in the preceding discussions how chemically slight the changes in a protein may be which can alter materially its antigenic nature, and it is a logical deduction that different organs of the same animal might contain antigenic constituents qualitatively different from the general serum protein. There are undoubtedly in many organs protein complexes which are peculiar to them and not present in other organs, and it would be reasonable to expect therefore that immunization with separate organ substances would lead to the pro- duction of sera of specific precipitating power for the protein of that particular kind of organ. This is not ordinarily obtainable, how- THE PHENOMENON OF PRECIPITATION ever, because it has been impossible to isolate from organs their pe- culiar, characteristic proteins, and immunization of animals with organ extracts or solutions has necessarily implied the injection of much blood protein and other albuminous material of a character general to many organs of the animal, i. e., to the species. These quantitatively overshadow the organ-specific substances which may be present, and give rise, therefore, to a "species" precipitin. That "organ specificity," however, is a fact has been shown by the experi- ments of Uhlenhuth with the protein of the crystalline lens of the eye. Immunization with this substance induces a precipitin which does not react with the serum of the animal from which the lens was taken, but does react, not only with the crystalline lens proteins of this spe- cies of animal, but also with crystalline lens proteins in general, though taken from another animal species. Analogous to this are the experiments of von Dungern and others upon the protein derived from the testicle. In both of these cases, as well as in other less sharply defined examples, the specificity is attached, not to the species of animal, but rather to the nature of the organ from which the particular protein is derived. These facts — first ascertained by means of the precipitin reaction — have been recently confirmed by means of the reaction of anaphylaxis by Uhlenhuth and Haendel, and by Kraus, Doerr, and Sohma. (See chapter on Anaphylaxis.) They have been discussed, moreover, in connection with the problem of spe- cificity in general. Biologically they probably signify that, although there are fun- damental species differences between the general body proteins of various animals, there are still, in certain highly specialized organs, varieties of protein which, possibly because of functional exigencies, have developed similar chemical characteristics. These have been determinable by our present methods, however, only for organs like the lens, the testicle, and the placenta from which the organ-specific protein can be gotten in a relatively pure state. The pathological importance of these phenomena lies in the fact that, although guinea pig serum injected into a guinea pig will not give rise to antibodies, lens protein apparently will do so — an observation which opens the possibility of autocytotoxins. The significance of this is indicated in such investigations as those of Homer,41 who, using the complement- fixation technique to determine antibody, found that the serum of adult human beings possessed antibodies for their own lens protein, but that such antibodies were absent in the sera of children. The study of agglutination and that of precipitation reveal, throughout, a close similarity between the two reactions, and indeed in physical principles they are probably the same, although the one 41R6mer. Klin. Monatsbl f. Augenheilkunde, Sept., 1906. Ref. from "Weichhardt's Jahresber.," Vol. 2, 1906, p. 348. 264 INFECTION AND RESISTANCE (agglutination) consists in the flocculation of large particles in sus- pension— the bacteria — while in the other the precipitation is one of smaller units — the precipitable colloidal particles of the protein solutions. This phase of the subject will be more thoroughly dis- cussed directly. Meanwhile, it is noticeable also that, even without drawing the physical parallel between the two reactions, there is much in the behavior of the antibodies — the agglutinins and the precipitins as conceived by Ehrlich, which led him and his school to attribute to them a similar receptor structure. Like the agglutinins, the pre- cipitins are not inactivated by 56° C., but when once rendered in- effectual by higher temperatures (70° C. or over) they can no longer be reactivated by the addition of fresh normal serum. For this reason chiefly Ehrlich has conceived that both agglutinins and pre- cipitins are "haptines" of the second order. CELL OR ^ law cat PRECIPITIN HflPTINE.ZH-° ORDER SCHEMATIC KEPRESENTATION OF EHRLICH 's VIEWS ON THE STRUCTURE OF CIPITINS. Ehrlich assumes that when dissolved protein substances — ordi- narily suitable for body nutrition — are injected into animals, they become anchored to the cells by such receptors of the second order. When overproduction occurs in response to repeated stimulation of the cells by consecutive injections (see Side-Chain Theory), these haptines of the second order circulate as agglutinins or pre- cipitins. Since they act without the apparent cooperation of alexin, he supposes that they carry within themselves the "zymophore," or ferment groups, by means of which the agglutination or coagulation is accomplished. It is this zymophore group which, it is assumed, accomplishes the digestion of the foreign protein before its assimila- tion, when these receptors are still parts of the living cell. Thus the conception of precipitins is identical with that formu- lated by the same school concerning the agglutinins, and the deduc- tions from these premises have been essentially similar. Thus, anal- ogous to the conditions prevailing in agglutination, Pick,42 and Kraus and v. Pirquet 43 have shown that when precipitating serum is inactivated by heat, and then is added to bacterial filtrates, it will 42 Pick. "Hofmeister's Beitrage," Vol. 1, 1902. 43 Kraus and von Pirquet. Centralbl f. Bakt., Vol. 32, 1902. THE PHENOMENON OF PRECIPITATION 265 prevent their subsequent precipitation by active precipitin. An illustration of this is found in the following protocol taken from the paper by Kraus and v. Pirquet (loc. cit., p. 69). (a) 5 c. c. cholera filtrate + 0.5 c. c. inactiv. (60°) cholera serum = no precipitate after 10 hours at 37° C. After 10 hours add 0.5 c. c. active cholera serum = no precipitate. (b) Omitted. (c) Omitted. (d) 5 c. c. cholera filtrate + 0.5 c. c. active cholera serum = after 10 hrs. typical precipitate. From this it was concluded that heat may destroy the zymophore or coagulating group of precipitins, leading to the formation of "precipitinoids" which, like agglutinoids, may have a higher affinity for the antigen than is possessed by the uninjured antibody. Subsequently there were opposed to these views the physical in- terpretations which have been outlined sufficiently in the section on Agglutination (see p. 240). In the case of precipitation the anal- ogy between colloidal reactions and the serum phenomena is fully as striking as in the former, an analogy in the delineation of which the first credit belongs to Landsteiner,44 45 and important further contri- butions have been made by Neisser and Friedemann, Forges, Gen- / gou, and a number of others. As in agglutination and colloidal flocV dilation, the presence of salts (electrolytes) fundamentally influ- ences the occurrence of precipitin reactions; and in both colloidal and precipitin reactions the relative concentration of the reacting bodies is paramount in determining whether or not precipitation takes place. In this connection the most frequently observed inhibi- tion occurring in serum precipitations is that which is caused by an excess of antigen. An example of this is as follows : Sheep serum 0.5 c. c. Antisheep rabbit serum Precipitate 1:10 4- 0.5 c.c. — 1:50 + 0.5 c.c. =b 1:100 4- 0.5 c.c. + + 1:500 4- 0.5 c.c. 4- + 4- 1:1,000 4- 0.5 c.c. ++ 1:5,000 4- 0.5 c.c. 4- This is entirely analogous to the inhibition which may occur when, let us say, a weak gelatin solution is added to a colloidal sus- pension of arsenic trisulphid ; or blood serum is added to mastic or arsenic suspensions. In both cases inhibition zones appear which 44 Landsteiner and Jagic. Munch, med. Woch., No. ' 18, 1903 ; No. 27, 1904; Wien. kl. Woch., No. 3, 1904. 45 Landsteiner and Stankovic. Centralbl. f. Bakt., Vols. 41 and 42, 1906. 266 INFECTION AND RESISTANCE show that the relative quantities of the two reacting bodies are quite as significant as their chemical or physical constitution in determin- ing the occurrence of flocculation. This, according to Bechold, Bil- litzer,46 and others depends upon the fact that the reason for floccu- lation is one of electrical charge. One hydrosol — say arsenic tri- / sulphid — can be flocculated by the oppositely charged colloidal ahu / minium hydroxid, but this will occur only when the quantitative^ relations are properly adjusted. If one or the other is in excess, no flocculation may occur, and, if subjected to a direct current, both colloids, though ordinarily wandering in opposite directions, will now wander in that of the one which is now present in the largest amount. We will not elaborate here upon the causes for this, since they have been indicated in the section on Agglutinins, and are set forth more accurately by Prof. Young in the special chapter on Col- loids. This effect of quantitative proportions would explain not only the absence of precipitation in the presence of too much antigen, but also the converse phenomenon, already mentioned, that precipitation may be inhibited when the precipitin is in excess. The fact that heated precipitating serum when added to its an- tigen not only does not cause flocculation, but may even prevent sub- sequent precipitation by active precipitin, also finds its analogy in colloidal reactions in the so-called protective colloids. Thus arsenic trisulphid may be protected from precipitation by gelatin, if a small amount of gum arabic is added, and the analogy has been brought even closer by Forges,47 who showed that heated serum will protect mastic suspension from precipitation by normal serum. This obser- vation of Forges is so closely similar to the results obtained by Kraus and v. Pirquet and others on the inhibition of precipitation by heated precipitating serum that it would seem, on first consideration, effec- tually to refute the conception of "precipitoids." However, it does not explain the specificity of such inhibition on the part of heated precipitating serum, as reported by Kraus and v. Pirquet, an observation which is one of the strongest arguments in favor of the derivation of the inhibiting factor from the specific precipitin (a precipitoid) ,48 In spite of the strong evidence in favor of the colloidal inter- pretations, such contrary evidence, brought forward by careful and 46 Billitzer. Cited from Bechold, "Die Kolloide, etc.," p. 79. 47 Forges. Chapter on "Colloids and Lipoids" in "Kraus u. Levaditi Handbuch," Vol. 1. 48 Although normal sera may gradually precipitate on standing, this takes place much more rapidly in precipitin-sera. The spontaneous precipi- tation of normal sera as well as of those under consideration is analogous to what Bechold and others call the "ageing" (altern) of colloidal suspen- sions, which, though originally stable, will eventually settle out, even in the presence of protective colloids. THE PHENOMENON OF PRECIPITATION 267 experienced workers, must be borne in mind and positive acceptance of the colloidal explanations, however attractive, must be withheld until much further investigation has been done. Another important and interesting phase of the study of precipi- tins is that associated with the occasional presence in the same serum of remnants of antigen and of precipitins which, though present side by side, do not unite to form precipitates. This condition is frequently seen in such sera as those produced by Fornet and Miiller 49 for rapid precipitin production for forensic work, a method in which the foreign serum is injected into rabbits in large amounts (2 to 10 c. c.), on consecutive days, and the animals are bled 6 to 8 days after the last injection. That such sera contain both antigen and antibody is shown by the fact that, though clear when taken, they will show precipitation not only when mixed with dilutions of the antigen, but also when added to homologous precipi- tating sera.50 This phenomenon has been noticed by Linossier and Lemoine,51 Eisenberg,52 Ascoli,53 and others, and has been extensively studied by von Dungern.54 Gay and Rusk55 have recently observed it in connection with the rapid method of precipitin production of Fornet and Miiller, and have noted that such sera, although containing both antigen and precipitin, do not possess complement-fixing properties. According to Uhlenhuth and Weidanz,56 the antigen may persist in the sera of protein-immunized animals, in demonstrable amounts, as long as fifteen days after the last injection, and it is constantly present during this period, but in progressively diminishing amounts. We are thus confronted by the apparently paradoxical phenom- enon of the presence in these sera, side by side, of an antigen and its homologous precipitin, incapable of reacting with each other, al- though each of them readily reacts with precipitin or antigen, re- spectively, when these are added from another source. Many attempts have been made to account for this. A number of observers, notably Eisenberg, have concluded from extensive an- 49 Fornet and Miiller. Zeitschr. f. biol Technik u. Methodik, Vol. 1, 1908. 50 For instance, a rabbit was injected on three consecutive days with sheep serum. It was bled on the fifth day after the last injection. The serum was clear when taken, but a precipitate was formed when it was added to sheep serum and also when it was added to serum from another rabbit similarly treated and containing sheep serum precipitin. 51 Linossier and Lemoine. C. E. de la Soc. de Biol., 54, 1902. 52 Eisenberg. Centralbl. f. Bakt., 34, 1903. 53 Ascoli. Munch, med. Woch., Vol. 49, No. 34, 1902. 54 Von Dungern. Centralbl. f. Bakt., 34, 1903. 55 Gay and Rusk. "Univ. of Cal. Public, in Pathology," Vol. 2, 1912. 56 Uhlenhuth and Weidanz. "Praktische Anleitung zur Ausfiihrung, etc.," Jena, 1909. 268 INFECTION AND RESISTANCE alyses of quantitative relationships, both of agglutinin and precipitin reactions, that these take place according to the laws of mass action. In consequence, in addition to the combined precipitin-antigen com- plex present in all mixtures of the two, there should also, be present free dissociated fractions of each, in amounts dependent upon rela- tive concentrations. This might explain conditions such as those described above. Yon Dungern, whose paper forms one of the most extensive studies of the phenomenon with which we are concerned, does not believe that precipitin reactions can follow the laws of mass action, and explains the simultaneous presence of precipitin and antigen in the same serum by assuming a multiplicity of precipitins. He believes that every proteid antigen contains a number of related partial an- tigens which give rise in the immunized animal each to a partial precipitin. In sera in which both antigen and precipitin are found side by side and free, he believes that the antigen is of a nature that has no affinity for the particular partial precipitin present with it. He says : "Auch hier handelt es sich nicht um zwei reaktionsf ahige Korper, deren Verbindung aus irgend welchen Griinden unterbleibt, sondern um Substanzen, welche keine Affinitat zu einander besitzen. Die betreffenden Kaninchen haben zu dieser Zeit noch nicht alle mb'glichen Teilprazipitine gebildet, sondern nur einzelne derselben. Diese zunachst produzierten, nur auf bestimmte Gruppen der prazi- pitablen Eiweisskorper passenden Partialprazipitine sind es, welche nach der Absattigung aller zur Yerfiigung stehenden zugehorigen Gruppen der prazipitablen Substanz in Serum nachweisbar werden. Daneben bleibt aber ein anderer Teil der prazipitablen Substanz, der keine Affinitat zu dem gebildeten Prazipitin bestizt, bestehen, solange bis ein anderes Partialprazipitin von den Kaninchenzellen geliefert wird, welches sich mit Gruppen der in Losung geliebenen Eiweisskorper vereinigen kann." Zinsser and Young57 have also studied these phenomena and have attempted to explain them on the basis of protective colloidal action. In considering the theories that have been advanced to ex- plain these occurrences, the conception of mass action as accounting for the simultaneous presence of the two reacting bodies in the same serum seemed entirely incompatible with our own observations and with those of Gay and Rusk, that these sera do not of themselves fix alexin. Were the conception of the manner of union of these two reagents, according to the laws of mass action, representative of the true state of affairs, it would be necessary to assume the pres- ence, in such sera, not only of the two reacting bodies free and disso- ciated, but also of a definite quantity of the united complex of the two, a state of equilibrium being established. If this were the case the sera should, in agreement with all experience on the phenomenon 57 Zinsser and Young. Jour, of Exp. Med., 1913, Yol. 17. THE PHENOMENON OF PRECIPITATION 269 of complement fixation, exert definite complement-binding power. Moreover, it has not been experimentally shown that colloidal sub- stances react in accordance with the laws of mass action as observed for simpler chemical substances. As regards the opinion of von Dungern, this seemed incom- patible with another occurrence, observed by many writers, namely, that such sera, although clear at first, eventually, after prolonged standing, do actually precipitate spontaneously; that is, the union of the precipitin and the precipitinogen does actually take place, but goes on with extreme slowness. Kow a notable and strange feature of this phenomenon is the fact that two such sera, both containing antigen and precipitin, but neither of them precipitating by itself, will precipitate each other when mixed. For this reason Uhlenhuth has advised against the use of mixtures of precipitin sera for forensic tests. For it is not unusual that precipitin sera, even when produced by the slow method, may contain traces of antigen, and this may lead to precipitate formation if such a serum is mixed with another homologous pre- cipitin and thereby simulate a positive forensic test. In seeking analogy for this serum phenomenon with the various colloidal suspensions, the problem consisted in protecting two mutually precipitating colloids by a third, and this in such propor- tions that the mixing of two such protected suspensions, each con- taining all three of the elements, would be followed by precipitation. This was obtained by the use of gum arabic, gelatin, and arsenic tri- sulphid. Thin emulsions of gelatin will precipitate arsenic tri- sulphid suspensions. Small amounts of gum arabic will act as a protective agent, preventing the precipitations. The amount of the protecting substance necessary to prevent precipitation in any one mixture varies apparently with every change in the relative proportions of the two. Thus a considerable number of mixtures of the three can be made which will remain stable for days, the actual and relative quantities of the three ingredients differing in each of the mixtures. When two such mix- tures are poured together, in many cases precipitation will result, varying in speed and completeness, according to the particular quan- titative relationship arrived at in the mixture. An example of such an experiment follows : Two solutions of colloidal arsenic sulphid were prepared, one containing 1 gm. per liter, the other containing 5 gm. per liter. With Kahlbaum's "Gold- ruck" gelatin a solution containing 1 gm. per liter was prepared. A solution of gum arabic was prepared which contained 10 gm. per liter, this being made stronger than the gelatin solution to avoid too great dilution in the final mixtures. The gelatin solution was prepared twenty-four hours before being used, as freshly prepared gelatin has but slight precipitating power for arsenic sulphid, this power appearing to increase greatly with the ageing of the solution. 270 INFECTION AND RESISTANCE For the purpose of demonstrating this analogy two protected solutions were prepared as follows : Solution 1. — This consisted of 2 drops of gum arabic, 2 c. c. of gelatin, and 5 c. c. of the weaker arsenic solution. Solution 2. — This consisted of 10 drops of gum arabic, 1 c. c. of gelatin, and about 4 c. c. of the stronger arsenic solution. In each case the arsenic sulphid was added until there were signs of increasing opalescence or turbidity, this being done in order that the two solutions should each be as little overprotected as possible. Portions of the two solutions were then mixed in equal propor- tions. In the course of a few minutes the mixture was noticeably more turbid than either of the original solutions. This turbidity continued to increase quite rapidly, and on the following morning after about sixteen hours of standing, the mixture was found to be completely flocculated out, while the original protected mixtures re- mained unprecipitated and showed about the same degree of opales- cence as on the preceding night. The same condition of affairs was found to have persisted after five days. On the fifth day the less concentrated of the clear protected suspension began to settle out, and was completely precipitated within twenty-four hours. The other remained clear for four days more, but on the ninth day it began to precipitate slightly, the precipitation remaining incom- plete. In these cases it appears, therefore, that a complete analogy to the observed conditions of the serum reactions has been found, and that all data observed in connection with sera in which antigen and precipitin are found side by side without reacting can be most simply explained on the conception of protective colloid action. Moreover, the chemical nature of the substances involved seems to add weight to this point of view. These relations have been gone into here at some length, since they seem to us to possess considerable theoretical and practical sig- nificance. For it may be that the presence of a protective colloid may, by inhibiting the union of antigen and precipitin within the body, protect the animal from intoxication during the early stages of immunization when antigen and antibody are present simulta- neously for longer or shorter periods. Were union between the two possible at such times in the circulation, an assumption necessitated both by the hypotheses of mass action and of multiplicity of precip- itins, there would probably be an absorption of complement by these complexes, with, as shown by Friedberger, a consequent formation of powerful toxic products. (See chapter on Anaphylaxis. ) It is not impossible by any means, therefore, that the injection of anti- gen in an animal in which such a balance has been established may THE PHENOMENON OF PRECIPITATION 271 lead to a sudden elimination of the colloidal protective action, union of the antigen and antibody, and, by the mechanism just outlined, anaphylactic shock. The fact, moreover, that mere heating will change the precipi- tating action, which certain sera have on inorganic colloids, to a protective one seems to show that this latter function may justly be associated with delicate physical or chemical alterations of animal sera. Furthermore, this point of view is strengthened by the fact that the mutual precipitation of sera, such as those described, takes place slowly, as does the mutual precipitation of two protected colloidal mixtures, in contradistinction to the more rapid precipitation which takes place when any of these sera is added to an antigen dilution, where the element of protection may be assumed to be practically eliminated by more extensively changed quantitative relations. CHAPTER XI PHAGOCYTOSIS EARLY investigations into the fate of bacteria within the infected animal body were largely carried out by pathological anatomists, and the observation of the presence of micro-organisms within the cells of the animal and human tissues was definitely made as early as 1870. Hayem,1 Klebs,2 Waldeyer,3 and others, saw leukocytes con- taining bacteria but failed to interpret this in the sense of possible protection. The process was regarded rather as a means of trans- portation of the bacteria through the infected body, or it was as- sumed that possibly the micro-organisms had entered these cells be- cause erf the favorable nutritive environment thus furnished. The first to suggest that such cell ingestion might represent a method of defence was Panum,4 who referred to it as a vague possi- bility. A similar iTut more convinced expression of this opinion was made in 1881, according to Metchnikoff,5 by Roser in explaining the resistance of certain lower animals and plants against bacteria. But Roser brought no experimental support for his contention, and little attention was paid to his assertion. The significance of cell ingestion as a mode of protection against bacterial invasion, therefore, was hardly more than a vague sugges- tion when Metchnikoff, who, though a zoologist, had become intensely interested in the problem of inflammation, began to experiment upon the cell reaction which followed the- introduction of foreign material, living or dead, into the larvae of certain starfishes (Bipinnaria). Pathologists, at this time, held complicated views of inflamma- tion which involved complex coordinated reactions of vascular and nervous systems, and MetchnikofFs primary purpose was to observe reactions to irritation in simple forms devoid of specialized vascular or nervous apparatus. He noted in these transparent, simple forms of life that the foreign particles were rapidly surrounded by masses of ameboid cells and reached a conclusion which, in his own words, is expressed as follows: 1 Hayem. C. E. de la Soc. Biol, 1870. 2 Klebs. Pathol. Anat. der Schusswinden, 1872. 3 Waldeyer. Arch. f. Gynekol, Vol. 3, 1872. 4 Panum. Virch. Arch., Vol. 60, 1874. 5 Metchnikoff. "L'Immunite dans les Maladies Infectieuses." PHAGOCYTOSIS "L'exsudat inflammatoire doit etre considere comme line reac- tion centre toutes sortes de lesions et Fexsudation est un phenomene plus primitif et plus ancien que le role du systeme nerveux et des vaisseaux dans I'lnflammation." 6 He compared the process of cell ingestion or phagocytosis of for- eign particles, as here observed, to that taking place in the most simple intracellular digestion which occurs in unicellular forms, a hereditary cell function now specialized in certain mesodermal cells, and passed on in the evolution of higher forms to other specialized cells. And indeed in animals of the most complex structure the leukocytes which carry on this phagocytic process may be considered as, in a way, representing a primitive form of cell, since they are only nucleated elements of the body which wander from place to place, and are anatomically independent of nervous control. In 1883, at the Naturalists' Congress in Odessa, Metchnikoff 7 first expressed his views and communicated the first of the splendid re- searches upon which our modern conception of phagocytosis is based. His earlier studies were carried out with a small crustacean, the daphnia, in which he studied the reaction which followed the intro- duction of yeast cells. He observed the struggle which ensued be- tween the ameboid leukocytes of the crustacean and the infecting agents and determined that complete enclosure of the yeast within the leukocytes assured protection to the daphnia, while a failure of this process, either from fortuitous causes or because of too large a quantity of the infecting agents, resulted in disease and rapid death. This early work of Metchnikoff forms the beginning of a long train of investigations to which we owe most of the basic facts we possess concerning the role of the phagocytic cells in the protection of the body against infection. Just as the various serum phenomena, of which we have spoken, have a general biological significance apart from their importance in relation to bacterial invasion, so the process of phagocytosis must be looked upon as an attribute of the animal and vegetable cell which has important physiological bearing entirely apart from infection. In fact, the ingestion of bacteria and other foreign particles by the leukocytes and other phagocytic cells of higher plants and ani- mals is entirely analogous to the intracellular digestive processes which take place, as the ordinary manner of nutrition, among the unicellular forms. Among the rhyzopods, in general, food is taken in by means of the ingestion of other smaller forms of life, bacteria, 6 Inflammatory exudation should be considered as a reaction against all sorts of injuries, and exudation is a phenomenon more primitive and ancient than are the parts played by nervous system and blood vessels in the process of inflammation. 7 Metchnikoff. Arb. a. d. zool. Inst., Wien, Vol. 5, 1883. 274 INFECTION AND RESISTANCE algae, etc. (or particles of dead organic matter), into the cell body of the protozob'n. These materials are gradually engulfed by the body of the ameba, which flows about them with its pseudopods, and within the cyto- plasm undergo gradual digestion. The process has been carefully studied by Mouton.8 In symbiotic cultures of amebae with colon bacilli on agar plates, the bacteria are taken up in large numbers and about them are formed small vacuoles. That the digestion takes place in a slightly acid medium with the vacuoles can be proved by adding a drop of neutral red to the hang-drop preparation of amebse and observing the brownish-red color taken by the materials in the vacuoles. Mouton was able to obtain a digestive ferment from the ameba?, by glycerin extraction, which exerted strong proteolytic action upon various albuminous substances, liquefied gelatin, and digested dead colon bacilli in vitro, acting best in slightly alkaline, but also in slightly acid, reactions. It is plain, therefore, that the most prim- itive form of digestion is an intracellular one carried on by ferments comparable in every way to the secreted digestive enzymes which accomplish the same purpose outside of the cells in higher animals. In essence, however, there is no fundamental difference physiolog- ically between intra- and extracellular digestions, and the intracellu- lar manner of assimilating solid nutritive particles may be retained in forms much higher in the scale of evolution than the rhyzopods. It has been studied by Metchnikoff and others in certain of the flat worms (Dendrocelum lacteum) in which typical phagocytosis is car- ried on by the cells of the intestinal mucosa. Many of these plan- aria obtain their nourishment by sucking the blood of higher ani- mals. Placed under a microscope after feeding, it may be seen that the foreign blood cells are rapidly taken up by the intestinal epithe- lial cells, which engulf them by means of pseudopodia not unlike those of the ameba. After ingestion, here, too, the blood cells are surrounded by vacuoles within which their gradual disintegration or digestion is accomplished. Similar intracellular digestion seems to be general among the crelenterates, and has been thoroughly studied by Metchnikoff in the actinia. Here the food particles are carried by the tentacles into the esophagus, and are taken up by the endo- dermal cells of the so-called "mesenteric filaments/' where they are digested by a trypsin-like enzyme. In these animals digestion is entirely intracellular, though the ingesting cells are the parts of a specialized tissue. In other forms, still higher in the scale, although there is persistence of intracellular digestion, the extracellular process begins to be developed. Thus in certain mollusca the solid food is taken into the intestinal canal, where it first undergoes a preliminary digestion by secreted intestinal juices. After it has 8 Mouton. C. E. de VAcad. des Sciences, Vol. 133, 1901. PHAGOCYTOSIS 275 been reduced to small amorphous particles in this way, these are seized by the ameboid cells, and intracellular digestion completes the process which has been begun extracellularly. As we study the process among higher animals, it appears that, among vertebrates, the intracellular methods of digestion have been, at least for normal metabolism, entirely displaced by the extracellu- lar as it occurs in the intestine, where solid particles are rendered completely amorphous, dissolved, and reduced to a diffusible condi- tion by the digestive juices before they are offered to the cells for utilization. However, the capacity for intracellular digestion is not entirely lost, and is retained of necessity in certain body cells. For were there not such an emergency arrangement the body would lack an available mechanism with which to meet such accidents as ex- travasations of blood, or the entrance of bacteria and other foreign solid particles into the tissues. It seems reasonable to classify both the phagocytic action of body cells and the formation of antibodies in the blood plasma, primarily as emergency devices for the diges- tion of foreign materials both formed and unformed which, under abnormal conditions, penetrate into the physiological interior of the body (blood stream or tissue spaces), and must be disposed of. In the lowest animals the single cell is called upon to perform all necessary functions. In the course of evolution, however, as the body becomes more and more a community of many cells, a division of labor takes place which is expressed morphologically in the differ- entiation of tissues and organs, and physiologically in the adaptation of individual tissue cells to the performance of specialized functions. Nevertheless, it is necessary, both for certain normal processes, as well as for provision against such complex emergencies as those mentioned, that certain cells of the complex community should retain the primitive abilities of the more independent cells of the lower forms. Thus, among many animals, the phagocytic action of cells performs definite services in the course of normal development. This is seen most markedly in some insects (diptera) in which the destruction of larval organs, useless to the adult animal, may be en- tirely accomplished by the action of phagocytic cells, and a similar process may accompany the transformation of the tadpole to the adult in many amphibia.9 In higher animals the removal of extravasations of blood is accompanied by a train of occurrences which is readily subjected to study.10 In such cases the leukocytes rapidly enter the area of extravasation and an engulfment of the blood cells occurs, followed by a process of digestion entirely analogous to the digestion of similar blood elements by the various forms of intestinal hem- ameba?. In the latter case it is a process of normal digestion, in the 9 See Henneguy. "Les Insectes," Paris, 1904, p. 677. 10 Langhans. Virchow's Archiv, Vol. 49, 1870. 276 INFECTION AND RESISTANCE former an emergency procedure carried out by virtue of the retained ancestral characteristics of the special phagocytic cells. The leukocytes, whose chief functions seem to be associated with such processes of intracellular digestion, may, therefore, be looked upon as cells retaining primitive characteristics for definite physio- logical purposes. We shall see, however, that, to meet exceptional conditions, the process of phagocytosis may be carried out also by many other cells which are associated ordinarily with functions en- tirely apart from this phenomenon. During normal life in higher animals, too, constant destruction of red blood cells by phagocytosis takes place in the spleen and liver, and is described by Dickson 1 1 as occurring in the bone marrow as well ; and similar phagocytosis of red cells is seen in the hemolymph nodes. It is claimed by Metchnikoff, furthermore, that many of the degenerative and retrogressive processes which take place in the human body are carried on by the mechanism of phagocytosis. The rapid return of the puerperal uterus to the normal state is explained in this way, and work by Helme 12 seems to show that there is an actual phagocytosis of the hyperplastic uterine musculature during this period. The atrophic changes of senility, too, are attributed by Metchnikoff1314 to the same processes. The involution of the ovaries v is accompanied by active phagocytosis of portions of this organ, and Metchnikoff claims further to have shown that the de- generation of the nervous system during old age is accomplished by the phagocytosis of nerve cells by phagocytic elements derived either from the leukocytes or the neuroglia, or from both.15 The whitening of the hair, both in human beings and in old animals (dogs), is simi- larly due, he claims, to phagocytosis of the pigment by cells which wander in from the root sheaths. It is, up to the present time, im- possible to determine the stimulus to which this phagocytosis is due. Since the subject is a very important one, many studies have been made to determine which cells of the body of higher animals can take in and digest foreign particles and to classify them according to this power. Metchnikoff has distinguished between the "motile" and "fixed" phagocytes, the former the leukocytes of the circulating blood, the latter certain connective tissue cells, endothelial cells, splenic pulp cells, and certain cellular elements of the lymph nodes, 11 Dickson. "The Bone Marrow/' Longmans, Green, London, 1908. 12 Helme. Transact. Roy. Soc. of Edinburgh, Vol. 35, 1889. Cited from Metchnikoff. 13 Matschinsky. Ann. de I'Inst. Past., Vol. 14, 1900. 14 Metchnikoff. Ann. de I'Inst. Past., Vol. 15, 1901. 15 That the leukocytes are concerned in the destruction and resorption of dead tissues has been shown by Leber especially (Leber, "Die Entstehung der Entziindung," Leipzig, Engelmann, 1891). An accumulation of leukocytes about a bacterial focus or from any other stimulus is followed by tissue lysis due to leukocytic enzymes. PHAGOCYTOSIS 277 the neuroglia tissue, and, in fact, all phagocytic cells which are ordinarily confined to some definite localization in the body. Among phagocytic cells Metchnikoff further distinguishes between "micro- phages," by which he designates the polymorphonuclear leukocytes of the circulating blood and "macrophages." The ma- crophages include the fixed cells mentioned above, to- gether with the large mono- nuclear elements of the blood, in short, all phago- cytic cells except the micro- phages. Although no absolute functional differentiation is possible between the two, it is true, in a general way, that the microphages are concerned primarily with the phagocytosis of bacteria and especially of those POLYNUCLEAR LEUKOCYTES TAKING UP STA- which invade acutely, while PHYLOCOCCI. the macrophages are con- cerned especially with the resorption of cellular detritus, foreign bodies, and such bacteria as are more chronic in their activities, or are peculiarly insoluble. On the other hand, micro- phages may take up foreign particles and bacteria of all kinds under suitable condi- tions, and no sharp line can be drawn between the two varieties in this respect. Metchnikoff further be- lieves that the two classes of phagocytic cells differ in the nature of the protective substances they secrete and furnish in the blood plasma. This, however, is a problem concerning which there is much difference of opinion and which calls for KUPFER CELLS CONTAINING MALARIAL PIG- MENT. DlAGRAMMATICALLY DRAWN FROM A SECTION OF MALARIAL LIVER KINDLY FURNISHED BY DR. E. LAMBERT. in 9pnaT.fltp other place. The property of phago- 278 INFECTION AND RESISTANCE BAT LEPROSY BACILLI GROUPED IN THE REMAINS OF DEAD SPLEEN CELLS GROWING IN PLASMA. Drawn after illustration in Zinsser and Carey, Journal of the A. M. A., Vol. 58, 1912. cytosis is therefore an attribute of a considerable num- ber of different va- rieties of cells. In the circulating blood the polynu- clear leukocytes are the most actively motile and phago- cytic elements. The eosinophile cells may also take up foreign particles and bacteria, as may also the large lymphocytes. The small lymphocytes and mast cells are either entirely inac- tive in this respect, or, at least, possess phagocytic powers under ex- ceptional circumstances only. This does not mean, however, that these last-named cells may not accumulate at the point of invasion nor that they may not play an important part in the defence of the body. It is well-known, of course, that, in tuberculosis and a number of other con- ditions, the lymphocytes may form the majority of the cellular elements which accumulate at the site of the lesion. Among the fixed cells of the body it is probable that phagocytosis may be carried on by cells of many different origins, though the identification of cells in tissues is often a purely morphological prob- lem, and therefore fraught with many possibilities of error. Probably the most active fixed tissue cells are the endothelial cells of the blood vessels and those which line the serous cavities, the PHAGOCYTOSIS OF SENSITIZED PIGEON COR- PUSCLES BY ALVEOLAR CELLS OF LUNG. Drawing made after photomicrograph pub- lished by Briscoe, Journal of Path, and Bact., Vol. 12, 1908. PHAGOCYTOSIS 279 sinuses of the lymphnodes, and of the spleen. However, there are many other cells in addition to these which may be phagocytic. The writer, with Carey,16 has observed the active phagocytosis of leprosy bacilli by cells, probably of connective tissue origin, growing from plants of rat spleen in plasma. Phagocytosis by the cells lining the alveoli of the lungs has been observed by Briscoe.17 This author made the interesting observation that in cases of mild infection such cells can free the lungs of micro-organisms entirely without aid from the leukocytes of the circulating blood. It is these cells, too, which, in the ordinary conditions of life, take up the inhaled particles of dust and are, therefore, often spoken of as dust cells. The origin of the dust cells has often been the subject of controversy. In the embryo the alveoli of the lung, like the bronchi, are lined with columnar cells which are transformed into flattened epithelium as the alveoli ex- pand at the first inspirations after birth. These flattened cells, which constitute the alveolar or dust cells, are probably of epithelial origin, and as such are probably the only epithelial cells which act as phagocytes under ordinary conditions. Although no positive general statement is justified, we can yet say with reasonable accuracy that among the phagocytic fixed tissue cells the most important are the connective tissue and endothelial cells. The type of phagocy- tosis and the variety of cell which participates in it seem to depend to a great extent upon the nature of the substance which incites the process, or rather at which the process is aimed. Thus the large cells which, in tissues, take up the lep- rosy bacillus, those which are characteristic of tuber- culous foci, or those caused by blastomycetes, or by for- eign bodies, all have special appearances which are suf- ficiently characteristic to have diagnostic value. However, it is difficult to determine with certainty the origin of the cells which participate. The chemical nature of the substances taken up, moreover, often complicates the phagocytic process in such a way that different cellular elements are enlisted in succession in order that the ingested substances may be disposed of. Thus tubercle 16 Zinsser and Carey. Jour. A. M. A., March, 1912, Vol. 58. 17 Briscoe. Jour, of Path, and Bacter., Vol. 12, 1907. GIANT CELL IN TUBERCULOSIS. INFECTION AND RESISTANCE or leprosy bacilli which are injected into an animal may be at first taken up by polynuclear leukocytes or microphages, by which they may even be carried into the lymph channels and distributed, per- haps to the detriment of the host. But these cells, probably because they lack a lipolytic ferment by means of which the waxes of the acid-fast organisms can be digested, cannot destroy the bacteria, which are then attacked by other cellular elements at the site of their final deposit. In many such cases the further resolution of the foreign sub- stance is accomplished by an important type of phagocytosis which is characterized by the formation of the so-called giant cells. These cells are of varying appearance in different conditions and locations. Thus the giant cells which form about foreign bodies, such as the small cotton fibers occa- sionally left in wounds, or injected particles of paraf- fin or iron splinters, etc., are quite characteristic and distinct from the giant cells of tuberculous foci, or of rhinoscleroma, glanders, or leprosy. They are all large cells, containing often nu- merous nuclei which form either by the fusion of sev- eral cells, as claimed by Borrell,18 Hektoen,19 and others, or by the cleavage of the nuclei alone, with- out coincident divisions of the cytoplasm. Although it is, of Dr. course, impossible to decide definitely upon purely morphological grounds, the researches of Hektoen especially would lead one strongly to favor the former view. It is equally difficult to decide the origin of giant cells, and endothelial, connective tissue, and even leukocytic origin has been claimed for them. Yet in no case has it thus far been possi- ble to actually observe their formation by a method which could posi- tively decide this point. In order to gain a clear conception of the participation of phago- cytes in the response of the body to injury or invasion, it will be useful to follow out the process of inflammation as it occurs in the 18 Borrell. Ann. de I'Inst. Past., 7, 1893. 19 Hektoen. Jour. Exp. Med., 3, 1898, p. 21. FOREIGN BODY OF GIANT CELL. SECTION OF CORNEA AFTER EXPERIMENTAL INJECTION OF PARAFFIN. After preparation kindly furnished W. C. Clarke. PHAGOCYTOSIS 281 higher animals. Inflammation may be incited by a large number of agencies — chemical irritants, mechanical injury, or even by the in- troduction of inactive and isotonic substances such as broth or salt solution.20 21 Yet in these cases the response, though essentially similar in principle to that following invasion by bacteria, lacks certain features especially interesting in the present connection, and it will be most profitable for our purpose to consider in detail the result of infection with pathogenic micro-organisms. If an emulsion of pyogenic staphylococci is injected into an ani- mal subcutaneously the site of injection will soon become reddened and swollen and microscopic examination will show, within a few hours, a swelling and engorgement of the blood vessels. The injected cocci will be found to lie partly scattered in the tissue spaces, in part within polynuclear leukocytes and connective tissue cells which have begun to ingest them. The tissue spaces will be swollen and stretched by the exudation of blood serum from the ves- sels. This condi- tion will begin in from 4 to 6 hours after injection and increase during the next 24 hours in ex- tent and severity, according to the quantity and viru- lence of the cocci injected. The con- ditions which pre- cede the wandering of the p o 1 y m o r- phonuclear leuko- cytes out of the ves- sels have been care- fully studied in such thin tissues as the mesentery of a frog after injury by trauma or acid. Within the vessels of the affected area there is at first an acceleration of the blood stream, then a dilatation of the capillaries and a slowing of the current. Leukocytes may now be observed moving more slowly than the main stream, and keeping close to the periphery along the walls of the vessels. Here and there they seem interrupted in their movements and adhere to the vascular wall. A little later these cells appear to pass through the wall of the 20 See Adami. "Inflammation," Macmillan, London, 1909. 21 See Adami, loc. cit. DIAGRAMMATIC REPRESENTATION OF LEUKOCYTES WAN- DERING THROUGH CAPILLARY WALLS. Adapted from Ribbert, "Lehrbuch der Allgemeinen Pathologic," p. 337. 282 INFECTION AND RESISTANCE vessel by sending out pseudopodia which slowly penetrate it. Adami states that if, at this stage, the tissues be excised, fixed in osmic acid, and stained, leukocytes may be seen crowding the inner sur- face of the vessel in all stages of transition from its anterior to the lymph spaces on the outside. In the staphylococcus infection, after from 12 to 48 hours, there will be seen the results of an active and destructive struggle between the invading bacteria and the defending cells. In the center of the area of invasion tissue has been destroyed and disintegrated. Amid the necrotic detritus, closely packed, lie leukocytes and cocci and active phagocytosis has taken place. In some cases the intracellular bacteria appear swollen and disintegrating, in others the leukocyte itself, overcome by the larger number of bacteria it has taken in, becomes vacuolated, indefinite in outline, and apparently is being itself destroyed. The presence of blood serum, which is aiding in the destruction of bacteria both by its bactericidal powers and its reenforcement of the phagocytic process, renders this mass fluid or semi-fluid, and the whole mixture constitutes what is known as pus. Around the periphery cocci and leukocytes become more scattered and sparse, and bacteria, together with leukocytes, loaded with cocci, may be seen lying within large mononuclear cells (macrophages). Whether the process goes on to further extension or is eventually walled off into a distinct abscess by the formation of granulation tissue and new connective tissue depends upon the balance of forces between attacking agent and defensive factors. If we inject a similar emulsion of cocci into the pleural or peri- toneal cavity of an animal a process similar in principle may be observed. Normally the peritoneum contains a small amount of this serous fluid and a moderate number of white blood cells, chiefly lympho- cytes. When any substance, broth or salt solution, an aleuronat or a bacterial emulsion, is injected into the peritoneal cavity, there follows a brief period during which there is a diminution of the free cellular elements in the peritoneal fluid. At this time there is a clumping of cells in the folds of the omentum and mesentery, a transient stage of flight away from the point of injury. This, how- ever, is soon over. Within one to two hours an active immigration of leukocytes into the serous cavity occurs and if, during the next 12 to 24 hours small quantities of fluid are, from time to time, with- drawn with a capillary pipette, a rapid and constant increase of leukocytic elements, chiefly of the microphage or polfnuclear type, is observed. If the injected substance has been a sterile, harmless fluid, a gradual return to normal within 48 hours then ensues. If, however, we have injected bacteria, a struggle similar to the one described above takes place within the peritoneum, and active phagocytosis of the micro-organisms takes place. PHAGOCYTOSIS 283 Let us suppose that the injected bacteria have been small in quantity and moderate in virulence. In such a case a rapid phago- cytosis gradually rids the fluid of micro-organisms and within 24: hours after injection few, if any, free bacteria are visible. A little exudate taken at this time shows large numbers of micro- phages varyingly crowded with well-preserved and disintegrating bacteria. Some of the phagocytes, having literally taken up more than they can digest, are vacuolated and disintegrating, but, in gen- eral, the victory lies with the cells. A little later large mononuclear elements appear, and here and there will be seen to take up dead leukocytes together with ingested cocci. In this way gradually a cleaning out of the peritoneum takes place, the animal recovers, and the peritoneum returns to normal. Let us suppose, on the other hand, that the bacteria injected are in larger doses and of greater virulence. In such a case, after a period of active phagocytosis, there may be a gradual increase of bacteria over leukocytes. The phagocytic cells are found to be under- going degeneration in larger numbers, the free bacteria increase, and the impending death of the animal can often be foretold by the appearance of the exudate. Finally, the peritoneal fluid may con- sist chiefly of free and rapidly multiplying bacteria with a practical absence of phagocytic cells. In all of the processes so far as described the burden of the defence has fallen upon the microphages or polynuclear leukocytes, while the macrophages — endothelial and connective tissue cells — have taken a purely secondary part in the reaction, forming, to some extent, a second line of defence, or, more probably, taking part only in the final removal of degenerated and disintegrating combatants and tissue detritus. In order to obtain a complete conception of phagocytosis in its entire significance it will be necessary to consider a further example, namely, the process which takes place within tis- ^sues in the course of the efforts of macrophages to remove bacteria and other substances which, either because of their insolubility or for other unknown reasons, are refractory to the attacks of the mi- crophages. Since we are interested in this subject chiefly from the point of view of the defence against bacteria, we may illustrate this process best by the description of the reaction which takes place when tubercle bacilli become localized anywhere within the animal body. When tubercle bacilli are injected into the peritoneum they are actively taken up by the polynuclear leukocytes just as are other bacteria and many entirely inactive solid particles. A similar inges- tion by microphages may take place in the folds of the intestinal mucosa if tubercle bacilli are fed to guinea pigs. However, this preliminary phagocytosis is probably of but secondary significance in the combat of the body against tuberculosis, since it has still to be shown that polynuclear leukocytes are capable of digesting and 284 INFECTION AND RESISTANCE destroying acid-fast bacilli. Indeed, much evidence tends to show that the ingestion of tubercle bacilli by microphages may be a detri- ment to the host, since the bacilli by this means are carried through the lymphatics and variously distributed throughout the body. Poly- nuclear leukocyte extracts, though containing, as we shall see, pro- teolytic enzymes, do not, according to Tschernorutzky, contain any lipase, and it may well be that for this reason they are unable to attack the waxy substances which form an integral part of these or- ganisms. This is in keeping with the observations made by Terry in our laboratory, that rat leprosy bacilli may be kept within leu- kocytes for weeks without losing their acid-fast properties, whereas the same bacilli, as the writer and Gary found, were rapidly disin- tegrated in spleen cells growing in plasma. Moreover, it is well known that the estimation of tuberculo-opsonin contents of the sera of tuberculous patients has been peculiarly unsatisfactory in throw- ing light on the progress of the disease. It would seem, therefore, that in this disease, as well as in others caused by acid-fast organ- isms, the microphages play only an unimportant part in the defence of the body. On the other hand, when tubercle bacilli are deposited either in a lymphnode (through the vehicle of leukocytes) or in a capillary anywhere by the blood stream, a train of cellular changes is initiated in which the predominant part is played by the macrophages. The tubercle bacilli so deposited are rapidly surrounded by large mono- nuclear cells, probably endothelial in origin. Some of the micro- organisms may even be phagocyted and taken into these cells. These cells, spoken of as "epithelioid cells," surround the clump of bacteria in more or less concentric rings, and around these there is an accumu- lation of leukocytes, largely of the lymphocyte variety, with an ad- mixture of a very few microphages. Then by the fusion of endothe- lial cells, or possibly by division of the nuclei of some of these cells within the individual cell bodies, giant cells are formed which take up the bacilli. The further progress of the tubercle now greatly depends upon the balance of power. Often such a tubercle may heal, possibly because of complete intracellular digestion of the ba- cilli. On the other hand growth and multiplication may lead to a slow and dry necrosis of the center of such a mass of cells, leading to the condition spoken of as caseation. Epithelioid cells lose their outlines and staining properties, and go to pieces. The center of the lesion is a grumous mass, the periphery shows a few giant cells and connective tissue proliferation. It is always surprising to those who study these lesions for the first time how rarely they succeed in finding tubercle bacilli in microscopic sections prepared from such tubercles by the ordinary Ziehl-Neelsen method of staining. Repeated and careful examina- tion of such material may fail to reveal any acid-fast organisms, PHAGOCYTOSIS 285 though inoculation into guinea pigs is nevertheless successful, pro- ducing typical tuberculosis. Much 22 has studied this peculiar state of affairs particularly and has shown that, although such lesions may show no tubercle bacilli by the Ziehl-Neelsen carbol-fuchsin method, staining by a modified Gram technique will reveal numerous Gram- positive rods and granules which have lost their acid-fast properties. This, too, if true, and the evidence is very much in its favor, would point to an ability of the macrophages to digest the waxy substance of the tubercle and other acid-fast bacilli, a property not possessed by the microphages. It may, of course, mean on the other hand that the tubercle bacilli in the lesion have not developed the waxy condi- tion. CHEMOTAXIS AND LEUKOCYTOSIS The part played by the phagocytic cell in the defence of the body against the entrance of bacteria and other foreign substances consists, then, of two functionally different phases. The first is an active motion of the cells toward the point attacked, and their accumulation about the noxious agent, the second consists in the act of ingestion itself. The motion of the leukocytes toward the invading substances indicates a sensibility on the part of the cell to changes in its environ- ment incited by the foreign agent, and since the stimuli most likely to reach the leukocytes and bring about this alteration in the direc- tion of their movements are chemical in nature, the phenomenon is spoken of as "chemotaxis." This term was borrowed from Pfeffer,23 who studied similar phenomena in connection with many freely motile plant cells, spermatozoa, and bacteria. Since the change of direction brought about in a moving cell by such influences may be such as either to attract or to repel, the term "positive chemotaxis" is used to designate the former and that of "negative chemotaxis" the latter. The property of chemotaxis is of vital interest in the present connection, since, whatever may be our opinion regarding the relative values of phagocytosis and serum protection in immunity, the great importance of the phagocytic process cannot be questioned, and any agency which repels the approach of the phagocytes must be a detri- ment, while any factor which attracts them is, of necessity, a power- ful means of defence. In the investigations upon the nature of infec- tious diseases attention has been concentrated upon the phenomenon of phagocytosis, and the relations governing the act of ingestion have been very thoroughly studied. The details of the chemotactic phenomenon, however, though of equal importance, are much more 22 Much. "Beitrage zur Klinik der Tuberk.," Vol. 8, 1907, Hft. 1 and 4. 23 Pfeffer. "Untersuch. a. d. Botan. Inst. Tubingen/' Vols. 1 and 2, 1884 and 1888. 286 INFECTION AND RESISTANCE obscure. A large part of our sparse knowledge in this connection,, moreover, has been gained by studies not related to infection. The stimuli which determine the motion of cells are, of course, not necessarily chemical, and extensive studies have been made upon the effect of light waves in this connection. Although these inves- tigations are of great biological importance, they have little direct bearing upon the problems of tropism as related to bacteria and leu- kocytes and cannot therefore be considered here. Some of the earlier researches upon chemotaxis were those made- by Stahl 24 upon the slime-molds or myxomycetes. These organisms, possess the power of ameboid motion, and were observed by Stahl ta move toward or away from any given region, according to the nature of the substances with which they came in contact, Pfeffer sub- jected this phenomenon to closer analysis. Working with the sperma- tozoa of ferns, swarm spores, bacteria and infusoria, he elaborated an ingenious technique by means of which he was enabled to de- termine directly the negative or positive chemotactic properties of various substances in solution upon these motile forms. His tech- nique was exceedingly simple. Capillary glass tubes, about 8 to 10 mm. long and 0.1 mm. in diameter, were sealed at one end in the- flame, and then dropped into a watch-glass. The solution which waa to be tested was poured over the tubes and the watch-glass then, placed under the bell of an air-pump. When the air was evacuated and pressure reduced the tubes became partly filled up with the liquid. They were then removed, washed in water, and placed under a cover slip under which a preparation of the motile cells was swim- ming. Positive chemotaxis was indicated by entrance of the cells, into the tubes, negative, by their refusal to enter. Failure of the solution to exert any chemotactic influence resulted in their moving- into and out of the tubes indiscriminately.25 By this technique a large number of interesting observations were made which threw much light upon the causes underlying the move- ments of plant cells. For instance, in investigating the spermatozoa of the ferns it was found that they were attracted strongly by malic acid and its salts, while no other substance investigated approached these compounds in the intensity of positively chemotactic stimula- tion. From this Pfeffer concludes that the bursting of the fern archegonia is accompanied by the liberation of malic acid, this at- tracting the male to the female cell. Similar experiments have been carried out since then by numer- ous naturalists, among them Buller,26 Lidforss,27 and Jennings,281 24 Stahl. Botanische Zeitung, 1884. 25 Buller. Annals of Botany, Vol. 16, No. 56, 1900. 26 Buller. Loc. cit. 27 Lidforss. "Jahrbiicher f. wissensch. Botanik," 41, 1904. 28 Jennings. "Behavior of Lower Organisms," Columbia Univ. Press.,, Macmillan, 1906. PHAGOCYTOSIS 287 and it has been found that in addition to malic acid compounds many other substances, organic and inorganic, occurring in plant cells and cell-sap exert positive chemotactic power. Lidforss has shown, for instance, that calcium chlorid in 0.1 per cent, solution may strongly attract plant spermatozoids (equisetum — horsetail). When the solu- tion is concentrated to 1 per cent., attraction is still exerted, but the spermatozoids immediately lose their motility upon entrance into the fluid. The same worker has shown that a substance which is positively chemotactic for one variety of plant cell may be negatively chemo- tactic for another, showing a certain selective variation which should be of great biological importance. Thus capillaries with a 1 per cent, solution of potassium malate actively attracted the spermatozoids of marchantia (a liverwort), while not a single spermatozoid of equi- setum would enter these tubes. Low 29 has applied these methods of study to the investigation of the chemotaxis of mammalian sperma- tozoa and found that these cells were actively attracted by weakly alkaline solutions. Studies upon the factors determining the movement of bacteria and amebse toward some substances and away from others have been numerous, and are valuable for the understanding of leukocytic chemotaxis, because they have led to the formulation of a number of important general theories. The fact that the motions of bacteria in suspensions are, to a certain extent, determined by the negative electrical charge which they all carry in neutral media, has been touched upon in the section on agglutination. Attempts on the part of Young and the writer to determine whether the attraction of leukocytes toward bacteria might be due to the carrying of an elec- tropositive charge by the white cells have met with no result, owing so far to the failure to elaborate a reliable technique. However, this thought is not an impossible one and should be borne in mind. That certain bacteria will wander actively toward a source of oxygen was shown by Engelmann' s 30 classical experiment in which a diatom, half in the shade and half in the light, was surrounded by an emulsion of bacteria, and these were seen to collect about the lighted half only, where oxygen was being liberated by virtue of the chloro- phyll. The extreme delicacy of chemotactic reactions is illustrated in these experiments in that Engelmann calculated that the bacteria reacted to one one-hundred billionth of a milligram of oxygen. The selective reaction of bacteria to various chemical substances, further- more, has been shown by allowing different solutions to diffuse into bacterial emulsions from capillary tubes, and by observing attraction or repulsion from the point of contact. The chemotaxis of leukocytes has opposed more difficulties to 29 Low. Sitzungs Berichte kais. Akad. d. Wiss., Wien, Vol. 3, Abf. 3. 30 Engelmann. Arch. f. d. ges. Physiol., Vol. 57, p. 375. 288 INFECTION AND RESISTANCE direct study, since the conditions within the living body are subject to a large number of modifying factors, and experiments upon the isolated cells, in vitro, even under conditions of the most careful technique, are fraught with much unavoidable injury to the cells. However, enough has been learned to indicate that these cells are subject to the phenomena of chemotaxis or tropism just as are inde- pendent unicellular forms, and that they may be attracted or re- pelled by a variety of organic and inorganic substances. Leber 31 was one of the first to study this in his work upon inflammation. He found that leukocytes were actively attracted by powdered cop- per and mercury compounds, but not by powdered gold or iron. He also observed that dead bacteria exerted a similar positive chemo- tactic influence, and Buchner 32 later succeeded in extracting sub- stances from various bacteria which possessed similar properties. It appears, from these and other investigations, that the power of stim- ulating positive chemotaxis is a general property of bacterial pro- teins, equally evident in bacterial extracts, dead bacteria, or the liv- ing organisms. It is likely, therefore, that the attraction of leu- kocytes toward the point of bacterial invasion is, in part at least, due to the properties of the bacterial proteins themselves. That this, however, is not the whole story is evident from the work of Massart and Bordet,33 who showed that the products of cell destruction and disintegration possess similar positively chemotactic properties. This is true not only of the products of disintegrated tissue cells, but of those of the destroyed leukocytes themselves. Thus it appears that when any injury of tissue takes place, a stimulus which attracts leukocytes results, even when the injury is not accompanied by bac- terial invasion. This would explain the participation of leukocytes in reactions to injury, and in inflammations not of bacterial origin, and their local accumulation following the injection of insoluble inorganic substances. When bacteria are actually present, however, the added stimulus due to the diffusion of bacterial proteins probably increases the process to a degree often sufficient to meet the added requirements for protection. Following this, both the destruction of tissues, of bacteria, and of leukocytes may together exert a cumulative chemo- tactic power which continues the process proportionately with the extent of the lesion. It is of the utmost importance, therefore, to ascertain whether or not any substances derived from bacteria may, under any circum- stances, exert a repellent or negatively chemotactic power. If we infect an animal intraperitoneally with virulent bacteria, in doses 31 Leber. Fortschr. der Med., 1888; also "Die Entstehung der Ent- ziindiing," Engelmann, Leipzig, 1891. 32 Buchner. Berl klin. Woch., Vol. 27, No. 30, 1890. 33 Massart and Bordet. Ann. de I'Inst. Past., Vol. 5, 1891. PHAGOCYTOSIS 289 sufficient to lead to death, and examine the peritoneal exudate just before the lethal outcome, we may observe that leukocytes are gradu- ally disappearing, and that finally but a few will be present and the fluid will be swimming with free micro-organisms. In the same way it is well known that the diminution of leukocytes in the circulating blood — or even the failure of these cells to increase in the circulation in the course of such diseases as pneumonia, or general infections with staphylococci or streptococci — is seriously prognostic of fatal outcome. The conditions here observed point strongly to the ex- istence of substances of negative chemotactic influence which protect the bacteria, not from phagocytosis itself, but from that necessary forerunner of phagocytosis, the approach of the leukocyte. It is necessary to draw this distinction since these phenomena are not merely, as often believed, "antiopsonic," but in truth largely "anti- chemotactic." It is true that Kanthack,34 and more especially Werigo,35 have denied the existence of negatively chemotactic bac- terial products, the latter basing his assertion upon the observation that active phagocytosis occurs in the lungs, liver, and spleen of animals dying of infection with virulent germs. However, the* argu- ments of these authors are not conclusive and the mass of experi- mental and clinical evidence which points to a direct failure of leukocyte accumulation in the presence of virulent bacteria in the animal body would alone suffice to render such conclusions unlikely. Moreover, strong evidence in favor of the existence of negatively chemotactic influences, is brought by the extensive experiments of Bail upon the so-called aggressins, discussed in another place, and such observations as those of Yaillard and Vincent 36 and Vaillard and Rouget,37 which showed that the injection of a little tetanus toxin together with tetanus spores would prevent the ingestion of the spores by leukocytes, and thereby furnish an opportunity for germi- nation and consequent fatal toxemia. Similar observations have been made by Besson 38 in the case of the bacillus of malignant edema by the use of the original technique of Pfeiffer. Capillary tubes containing the toxin remained free of leukocytes after subcutaneous introduction into guinea pigs, while similar tubes containing the culture medium alone, or the bacilli and their spores, attracted leukocytes in considerable numbers. It is possible, of course, to interpret such phenomena as due to a failure of positive chemotaxis rather than to an active negative chemotaxis. Although the phenomena of chemotaxis are most easily studied 3* Kanthack. Quoted from Adami, loc. cit. 35 Werigo. Ann. de I'Inst. Past., Vol. 8, 1894. 36 Vaillard and Vincent. Ann. de I'Inst. Past., Vol. 5, 1891. 37 Vaillard and Rouget. Ann. de I'Inst. Past., Vol. 6, 1892. 38 Besson. Ann. de I'Inst. Past., Vol. 9, 1895. 290 INFECTION AND RESISTANCE in extravascular inflammatory changes, there is none the less a regular and apparently purposeful attraction or repulsion of leuko- cytes evident in the circulating blood during infectious diseases. That infection of the body with many micro-organisms results in the increase of leukocytes, and that in others there is either no increase or even a decrease, is too well known and too generally applied in diagnosis and prognosis to warrant our giving up much space to a review of the facts. Nevertheless, the causes which lead to a leuko- cytosis in the one case, a leukopenia in the other, are still very obscure and deserve discussion. In the first place it is by no means certain whether a leuko- cytosis signifies an active discharge of new leukocytes from the bone marrow or whether it means simply an altered distribution in that the phagocytes accumulated in the lymphatic and other organs are attracted by chemotaxis into the peripheral circulation. Studies of the bone marrow during infection as well as the occasional appear- ance of myelocytes and other cells ordinarily found only in the bone marrow during health would point toward a participation of active bone-marrow hyperplasia in the increase of peripheral leukocytes. There is no good reason to doubt, moreover, that a chemotactic stimu- lus exercised in the circulation should withdraw leukocytes from any place of accumulation to the circulation. Probably both proc- esses take part. When bacteria are injected into the circulation of an animal there is, at first, a moderate diminution of the leukocytes just as there is after injection of bacteria or other substances into the peritoneum. This is soon followed in most cases by a rapid and progressive increase, in which, whenever the leukocytosis is one of considerable degree, the polynuclear leukocytes preponderate. The extensive clinical study of the white cells in infectious disease of the human being give us more material for reasoning in this respect than we have available from animal experiment. Infection with invasive bacteria such as the pneumococcus (and Neufeld and others, have shown that most lobar pneumonias are accompanied by pneu- mococcus bacteriemia), streptococci, staphylococci, and others is always accompanied by an increase of the leukocytes, while, in typhoid fever, influenzal infection, tuberculosis, and a number of other infections, the leukocytes do not increase and may even de- crease. How are we going to account for this ? That all these bac- teria contain a substance which is positive in its chemotactic effects is easily demonstrated by injecting them into the peritoneum and observing an accumulation of leukocytes and a consequent phago- cytosis, even in the cases of those organisms which do not call forth a leukocytosis in the blood of the diseased human being. Thus it has been our experience as well as that of others invariably to ob- serve the rapid and complete polynuclear phagocytosis of both leprosy bacilli and tubercle bacilli after injection of these micro- PHAGOCYTOSIS 291 organisms into the peritoneal cavities of guinea pigs. Yet a chronic tuberculous peritonitis or pleurisy is characterized usually by an exudate which contains but few polynuclears and relatively many lymphocytes. A final explanation of these conditions is not pos- sible at present. No adequate explanation for the selective accu- mulation of lymphocytes and the absence of polynuclear cells about tuberculous foci has yet been advanced. The absence of polynuclear leukocytosis may possibly be due to the great insolubility of these bacilli, in consequence of which little or no leukocytosis-stimulating substances are liberated. Pearce 39 has suggested a similar reason for the absence of poly- nuclear accumulations about chronic localized lesions of any kind in which tissue encapsulation may prevent the contact of the inciting agents with the body fluids and there is a consequently slow or slight production of such chemotactic stimulating materials. In typhoid fever, where the slight primary leukocytosis is rap- idly succeeded by a leukopenia with a relative lymphocytosis, the conditions are somewhat different. Here, as in some other infec- tions, as Friedberger and others have shown, we are dealing with a generalized infection by an organism which is easily subject to the action of alexin with consequent production of anaphylatoxin. (See chapter on Anaphylaxis.) This poison, it seems, exerts a nega- tive chemotaxis, and probably during the height of the disease, therefore, leads to the low leukocyte count observed. That this is at least likely seems to follow from the studies which have been made upon the nature of the typhoid poisons, and also from the observation of Gay and Claypole, that typhoid-immune rabbits react to the infection of typhoid bacilli with a rapid and powerful increase in the polynuclear leukocytes, whereas similar injections into the normal animal lead to leukopenia. If the supposition regarding tuberculosis, made above, is correct, it would follow that a sudden and considerable increase in the poly- nuclear leukocytes in a case of tuberculosis would indicate a dis- charge of organisms into the circulation and a tendency toward gen- eralization of the infection in this manner. (See Weigert's view of the manner in which tuberculosis may spread by the destruction of the wall of a vein by a localized lesion.) However, although specu- lation in the absence of experimental proof is justified, it must not be forgotten that the problems of selective chemotaxis are too ob- scure to permit of our laying much weight on any of these views. Gabritchewsky,40 who investigated this subject extensively, has classified various substances according to their positively, negatively, or neutral chemotactic activities. It is not necessary to recapitulate these, but it is interesting to note that he found that some substances 39 Pearce. Jour. A. M. A., Vol. 61, 1913. 40 Gabritchewsky. Ann. Past., Vol. 4, 1890. 292 INFECTION AND RESISTANCE which were positively chemotactic in certain concentrations became neutral or even negative when the concentration was altered. We have seen that the action of the leukocytes in moving toward some substances and away from others is entirely analogous to simi- lar phenomena occurring among lower, unicellular forms of life, and the explanations applied to the apparently conscious acts of the ameba, such as the motion toward and the engulfment of food, have been applied to the activities of the leukocytes as well. Many of the theories developed concerning the free living forms, however, have been easily excluded in the case of the leukocytes, because of the environment in which their activities are developed. Thus the many interesting reactions of paramecia and other organisms to light (heliotropism) have little bearing upon this subject, and the views based on the theories of orientation may be excluded on the ground of the symmetry of the normal leukocyte. The observations of Garrey,41 that indicate that it is the dissociated ions of various acids and bases which are responsible for the directive stimuli exerted upon certain flagellates, may yet result in throwing some light upon leukocytic movements, especially if we can come to accept the con- ceptions of ion-proteins upheld by Loeb 42 and his pupils. However, the facts concerning these phenomena, as well as the possibility, previously mentioned, of the opposite electrical charges carried by the leukocytes and the substances attracting them cannot be regarded at present as more than interesting thoughts. Of more than merely speculative interest, however, are the views of chemotaxis which are based upon the study of conditions of surface tension. In order to consider these properly it will be useful to review briefly the funda- mental principles governing these conditions. The molecules of any fluid are held together by mutual attraction due to the force generally spoken of as cohesion. This force is ex- erted by like molecules upon each other in solids more strongly than in liquids, and in gases less strongly. Since we are dealing in this connection with occurrences taking place in liquids, we will restrict our consideration to these. The force of cohesion is influenced in a number of ways. Thus, for instance, heat reduces it, and this is the cause that solids are converted into liquids and liquids into gases, provided of course that the heat brings about no chemical change. In large masses of fluids the force of gravitation over- comes that of cohesion and larger masses of liquid assume the shape of the containing vessel. In smaller masses the force of cohesion tends to bring about the spherical shape. This comes about in the following way: Within the interior of a drop of liquid all the molecules attract each other, and since the force of attraction is equal in all directions it neutralizes itself, and the molecules are 41 Garrey. Am. Jour, of P~hys., 3, 1900. 42 Loeb. Am. Jour, of Phys., 3, 1900. PHAGOCYTOSIS 293 uninfluenced by it, mobile and free. The molecules on the surface are in a different condition, however. They are subjected to the force of cohesion from within, but not from without, and are there- fore drawn strongly toward the center. The result is the same as though the drop were subjected to pressure from without and the surface layers were in a state of compression. There is in conse- quence a constant tendency of all the surface molecules to be drawn toward the center and a resulting tendency to a diminution of the surface area. It is as though the surface of such a drop were a thin, elastic membrane which tended to contract and diminish in size and surface. The force with which this takes place is spoken of as sur- face tension,43 and the energy underlying it is called, by Ostwald, surface energy. Since a drop of one fluid suspended in another with which it cannot mix is relieved of the disturbing factor of gravitation, its surface tension has the effect of contracting the small mass into a form which, for the given volume, will expose the smallest possible surface, and this is, of course, the sphere. It is for this reason that, if we shake up such systems as water and chloroform, or oil and water, the chloroform or the oil will be distributed through the water as small droplets. The degree of surface tension of any fluid is meas- urable by a number of reasonably accurate methods which may be found in any text-book of physics and which we need not consider here. It is of course dependent in each case upon the nature of the surrounding medium. We have taken into consideration above only the force which is exerted within the drop by the cohesion, that is, the attraction toward the center. This would be uninfluenced from without only in a vacuum. In nature the surface molecules, though forcibly drawn toward the center, are also affected from without by the attraction exerted by the molecules of the substances surrounding the drop. There is a constant balance, therefore, at any part of the surface of a drop of fluid between the cohesion tension from within and attractions from without. The resultant of the two forces de- termines the surface tension, which will be greater or less in inverse ratio to the attraction from without for any given drop, and a varia- tion of the external attraction at different points on the periphery of the drop will naturally influence' the shape of the drop. For a relief of attraction at one point would tend to permit that part of the sur- face to retract, and an increase in this attraction would tend to allow it to bulge, with the formation of a sort of pseudopod. In studying the importance of surface tension 44 in determining the motions of unicellular organisms a number of important attempts have been made to imitate cell motion by means of the suspension of various substances of strong cohesive properties in liquid media. The 43 Michaelis. "Dynamik der Oberflachen," Steinkopf, Dresden, 1909. 44 For a thorough discussion of this phenomenon see also Gideon Wells, "Chemical Pathology," Saunders, 1911. 294 INFECTION AND RESISTANCE idea was suggested by Quincke,45 and later by Biitschli,46 but has been most extensively studied by Rhumbler.47 The result has been the production of a number of "artificial amebse" which in almost all respects behave like the living organisms. Thus if a small mass of mercury is placed into a dish filled with water acidified with nitric acid, and a small crystal of dichromate of potassium is dropped near the mercury, the dichromate will dissolve and a yellow cloud will gradually diffuse from it toward the mercury. As soon as the yellow cloud touches this it will begin to show change of form and to elongate in the direction of the dichromate, often moving to- ward it. The motion of the quicksilver will resemble with con- siderable accuracy that of an ameba moving toward a particle of food or sending out pseudopodia. A more striking and coriiplete imitation is that obtained by Rhumbler when he placed a drop of clove oil into a mixture of alcohol and glycerin. The changes of sur- face tension produced upon the surface of the clove oil by the alcohol give rise to movements in the oil entirely analogous to those of mo- tile cells in favorable media. The similarity has been extended even to the processes of engulfment of the food as observed among amebse. Thus a drop of chloroform in water will flow about a particle of shellac and dissolve it. If a piece of glass coated with shellac is placed in contact with the drop it will engulf it, but will cast out the glass after the shellac coating has been dissolved away. The similarity between phenomena purely referable to surface tension and those taking place in the living cells is therefore very striking and has been clearly analyzed in regard to its bearing upon leukocytic chemotaxis by Gideon Wells in his "Chemical Path- ology." The chemotactic substances, diffusing to the leukocyte, will lower its surface tension on the side at which they come in contact. Pseudopodia will be thrown out on this side in consequence, and the leukocyte will move in this direction. The motion will be continued in this direction as long as the concentration of the chemotactic sub- stance, and therefore the diminution of surface tension is greater on this side than on other parts of the periphery, until a point is reached at which the chemotactic substance is equally diffused on all sides, and motion will cease. The actual engulfment may then occur or the nature and concentration of the chemotactic substance may be so great that injury is done to the leukocyte. Whether or not the purely physical explanation of chemotaxis tells the whole story it is of course not possible to decide. At any rate, it furnishes a rational basis for 45 Quincke. Quoted from H. G. Wells, "Chemical Pathology/' Saunders, 1907. 46 Butschli. "Untersuch. iiber mikroskopische Schaume und das Proto- plasma." Leipzig, 1892. See also H. G. Wells, loc. cit., pp. 220 et seq. 47 Rhumbler. Arch. f. Entwickelungs Mechanik, 1898. PHAGOCYTOSIS 295 the study of the phenomenon more promising than any of the others so far offered. It is true, on the other hand, that such a theory in no way ac- counts for the apparently selective positive chemotaxis which is exerted by different substances. Thus the preponderance of poly- nuclear leukocytes in foci and serous cavities containing organisms like staphylococci, meningococci, streptococci, and others is in con- trast to the lymphocytic accumulation in the pleural, subarachnoid, and peritoneal spaces infected with tubercle bacilli. Some writers have spoken, therefore, of active and passive leukocytosis according to whether or not the cells attracted seemed to possess ameboid motility. That surface tension phenomena alone do not account for this is clear. But it must be remembered that even tubercle bacilli, though eventually attracting few polynuclears and many lympho- cytes, will cause an active polynuclear accumulation in the perito- neum and pleura when first injected, and are actively phagocyted. Later when the lesion is established and the bacilli are lodged in the tissues the polynuclears give way to the lymphocytes, which even then never accumulate in such proportion as do the microphages in acute suppurative lesions. It may well be that the chemotaxis origi- nally attracting the polymorphonuclear leukocytes is the same in every case, but that a continued irritant, especially one well sur- rounded by tissue elements as are the organisms within the tubercles, may cease to exact any chemotactic influence, the accumulation of in- active lymphocytes possibly being due to a progressive death of these cells carried into the neighborhood of the lesion by the normal circu- lation of the lymph. CHAPTER XII THE EELATION OF THE LEUKOCYTES AND OF PHAGOCYTOSIS TO IMMUNITY IN Metchnikoff's earliest work upon the daphnia or water flea he observed clearly that there was a direct relation between the de- gree of phagocytosis and the outcome of the infection. When phagocytosis of the invading yeasts was energetic and complete the daphnia recovered. When the yeast cells penetrated the intestinal wall of the daphnia in large numbers, and were enabled to multiply before the phagocytic cells could accumulate in sufficient numbers to engulf them, then the body of the daphnia was soon swamped with the parasites and death ensued. This simple observation fostered the thought that the basic prin- ciple underlying all processes of immunity was represented in this struggle between the invading bacteria and the phagocytic cells. To the activity of the latter, entirely, he attributed the power of resistance. In support of this contention Metchnikoff and his pupils have marshaled many facts, most of which are set forth in his classical work "L'Immunite dans les Maladies Infectieuses." It will be manifestly impossible here to do more than outline the plan of study which these investigations have followed and the conclusions to which they gave just foundation. The original study upon the infectious disease of daphnia led to analogous experiments upon higher animals and, by the prolonged and patient investigations of Metchnikoff and his pupils, it was shown that, throughout the field of infectious disease, there was a striking parallelism between the resistance of the infected subject and the degree of phagocytosis which occurred. Earlier studies concern themselves chiefly with the natural im- munity possessed by many animals against certain infection. The infectious disease which at this time had been most thoroughly studied was anthrax, and Koch had shown that frogs and other cold- blooded animals were markedly resistant against this micro-organism. Taking advantage of this observation, Metchnikoff studied the phago- cytosis of anthrax bacilli in frogs and found that it took place rap- idly and effectively, all of the injected bacilli being soon engulfed by the accumulating cells. Similarly, active phagocytosis of anthrax RELATION OF LEUKOCYTES TO IMMUNITY 297 bacilli was demonstrated in such naturally resistant animals as dogs and chickens, while almost no cell ingestion occurred in delicately susceptible animals like guinea pigs and rabbits. Rats, on the other hand, more resistant to anthrax than guinea pigs, less so than dogs, showed a degree of phagocytosis intermediate between that observed in the cases of the other animals mentioned above. And yet, in these more susceptible animals, the normal bactericidal action of the blood upon anthrax bacilli, though never extreme, was often more marked than that of the naturally immune animals mentioned above. It is well known, for instance, that the serum of dogs possesses almost no bactericidal properties for anthrax bacilli,1 although tne animals are highly resistant to this infection, while the serum of rabbits is probably more strongly bactericidal for these bacilli than the serum of most other animals, and 'yet rabbits are extremely susceptible. That the lack of bactericidal powers of the serum is not always a sign of susceptibility on the part of the animal was shown as early as 1889 by Lubarsch. (We must remember, how- ever, that lack of bactericidal power does not necessarily mean lack of sensitizer. For bacteria may be sensitized without being killed extracellularly as can be shown by the alexin-fixation reaction.) The study of anthrax infections was a peculiarly fortunate choice of subject, since in this bacillus resistance to serum lysis is especially well marked and phagocytosis seems indeed to be the chief mode of bacterial destruction. Studies analogous to those originally made with anthrax, however, were subsequently carried on with streptococci, pneumococci, and staphylococci chiefly by Bordet,2 Marchand,3 and others; and results coinciding with those of Metchni- koff were obtained. In every case naturally resistant animals showed marked phagocytosis, and susceptible ones failed to show it to the same degree. It is a strong support of the same opinions, too, that Marchand' s studies, later extensively confirmed, showed that the more virulent and invading strains of streptococci, the less active is the phagocytosis — a converse, but equally conclusive, observation. Further support for this point of view is manifold and cannot be considered with anything like completeness. We may refer briefly, however, to the experiments of Vaillard, Vincent, and Rouget 4 with tetanus, and those of Leclainche and Vallee 5 with symptomatic anthrax, because they are especially valuable in illustrating the importance of phagocytosis in another class of infection. The pois- ons of these micro-organisms are extremely toxic for rabbits, and if 1 Petterson. Centralbl f. Bakt., 1, 39, 1905. 2 Bordet. Ann. de I'Inst. Past., Vol. 11, 1897. 3 Marchand. Archiv. de med. Exp., Vol. 10, 1898. 4 Vaillard, Vincent, and Rouget, Ann. de I'Inst. Past., Vols. 5, 6, 1891- 1892. 5 Leclainche and Vallee. .Ann. de I'Inst. Past., Vol. 14, 1900, 298 INFECTION AND RESISTANCE a small amount of culture material, together with agar, broth, or any foreign substance which may inhibit or divert phagocytosis from the spores, is injected into these animals rapid proliferation and death with toxemia result. If, on the other hand, the spores are carefully washed of foreign material and toxin rapid phagocytosis results and the animals recover. The parallelism which was followed out so extensively between natural immunity and phagocytosis was even more closely marked in the case of artificially acquired immunity. The first observations of this kind made by Metchnikoff, again on the subject of anthrax infection, were carried out by the active immunization of rabbits. The subcutaneous injection of virulent anthrax bacilli into normal rabbits is usually followed by a rapid growth of the bacteria, with much serous exudation but hardly any leukocytic accumulation. In immunized animals, on the other hand, the bacilli are taken up by hosts of phagocytes, just as this occurs in naturally resistant dogs or other animals. Similarly Bordet 6 has shown that cholera spirilla injected into the blood stream of cholera-immune animals are taken up by leukocytes even before they can be subjected to lysis by the circulating lytic antibodies. It would add little to clearness were we to multiply the examples in which it has been demonstrated that the acquisition of increased resistance is accompanied by enhancement of the phagocytic process. This statement may be regarded as an axiom, and indeed our later discussions of the opsonins and bacteriotropins will show clearly why such a state of affairs is to be expected. Taken by itself, however, it does not necessarily prove that the destruction of the invading germs is entirely due to the leukocytes. It might still be possible that the bacteria are injured or even killed by the antibacterial serum constituents before they can be taken up and carried away by the cellular elements; the phagocytes then would act only as scavengers for the removal of the dead bodies. Indeed, this opinion was long held by a number of the adherents of the purely humoral school. However, such a point of view is no ^nger tenable — espe- cially in the light of the later opsonin studies just referred to. Moreover, long before these more recent studies it was clear that bacteria may often grow within the leukocytes — finally destroying these — and that they may even remain fully virulent after ingestion. For, as Metchnikoff showed, if guinea pigs were injected with a little of the exudate formed after the injection of anthrax bacilli into immunized rabbits (an exudate in which there were no longer any extracellular bacteria because of energetic phagocytosis) death often resulted. It was clear, therefore, not only that the ingested bacteria were still alive, but that they were, at least in part, still fully virulent. 6 Bordet. Ann. de I'Inst. Past., Vol. 9, 1895. RELATION OF LEUKOCYTES TO IMMUNITY 299 A further method of investigation employed by Metchnikoff in his endeavors to prove his point consisted in the attempt to demon- strate that virulent bacteria could be protected from destruction in the bodies of resistant animals if the leukocytes could be held at bay. This resulted in a number of ingenious experiments, the most con- vincing of which is the one carried out with anthrax bacilli and frogs by Trapetznikoff.7 Anthrax spores were inclosed in little sacks of filter paper and these were introduced subcutaneously into frogs. In consequence the spores, bathed in the tissue fluids, but protected from phagocytosis, developed into the vegetative forms, multiplied, and remained virulent for several days. Taken up by the frog's phagocytes under ordinary conditions, they would rapidly have been taken up, digested, and destroyed. Here again it was shown that the body of fluids alone were unable to dispose of the bacteria and that the natural resistance of the frog was due entirely to phagocytic processes. Other experiments have been aimed at a general reduction of phagocytic activity by the use of narcotics. Thus, Cantacuzene 8 showed that animals treated with opium are very much more sus- ceptible to infection than are normal controls. And since opium markedly inhibits the activity of the white cells it may possibly be that these experiments furnish a further support for Metchnikoff' s opinion. At any rate, it is worth noting that, even though these experiments are not convincing in their assertion that the increased susceptibility was due entirely to the interference with the leuko- cytes, they indicate very definitely the inadvisability of using mor- phin and similar narcotics in infectious diseases. It is quite clear at any rate, then, that the process of phagocy- tosis increases in energy as immunity is acquired and, so far, Metch- nikofFs assertions are entirely upheld by later knowledge. In his contention that all properties upon which the resistance of the ani- mal against infection depends center directly or indirectly in the phagocyte, however, many subsequent amendments have been neces- sary, which will become self-evident in the following discussions of individual phases of the destruction of invading bacteria. We cannot at the present time attempt to correlate these extreme views of Metchnikoff with the equally extreme opinions of those in- vestigators who formerly attributed immunity entirely to the prop- erties of the body fluids, assigning to the cellular activities a merely secondary role. Many of the apparently opposed contentions have become reconciled, and we now realize that neither process alone tells the whole story, both being parts of the complicated correlated proc- esses which together constitute the mechanism of resistance. It was 7 Trapetznikoff. Ann. de VInst. Past., 1891. 8 Cantacuzene. Ann. de VInst. Past., Vol. 12, 1898. 800 INFECTION AND RESISTANCE indeed to the eager controversy between the two schools that we owe much of the clearness of conception which recent years have given. After the bacteria are taken up by phagocytes they undergo a gradual disintegration or dissolution comparable to that by which a particle of food is digested within the cell body of a rhizopod. With the exception of such particularly insoluble micro-organisms as the tubercle bacillus, the leprosy bacillus, blastomyces, and a few others, there is in all cases an eventual complete resolution of the bacterial body. As in amebse the digestion takes place often after the formation of digestive vacuoles, and by staining at this time with neutral red it may be demonstrated that the process goes on in a weakly acid environment. Metchnikoff naturally assumed, therefore, that the intracellular digestion of bacteria by microphages (polynuclear leukocytes), or of cellular elements, etc., by macrophages, was a process carried on most probably by enzymes, and that these enzymes were identical with the bactericidal bodies described as "alexin" and "sensitizer" or "amboceptor" in the blood stream. To follow without confusion the development of his ideas, however, it is necessary to bear in mind that much of his earlier work was done at a time when the discovery of the cooperation of two substances in bacteriolysis and hemolysis (the amboceptor and the complement) had not yet been made by Bordet, and when the bactericidal effect of normal serum was at- tributed entirely to a single substance — the alexin of Buchner. Buchner 9 himself had suggested that alexin was an enzyme-like body probably derived from the leukocytes. In his experiments Buchner had noticed that exudates, produced by intrapleural injections of aleuronat in rabbits and dogs, possessed a bactericidal value for Bacillus coli which exceeded the bactericidal power of the blood serum itself. The influence of active phagocytosis could be excluded by the fact that the leukocytes of the exudate had been killed by repeated freezing and thawing. Similar results were obtained by Hahn 10 with B. typliosus. Denys and Kaisin,11 working along similar lines, found that the pleural exudates of rabbits, obtained by the injection of dead staphy- lococci and freed ot cells by centrifugalization, were more highly bactericidal for staphylococci than the blood serum of the same ani- mals, but found also that the inactivated exudate could not be reacti- vated by the addition of leukocytes. Denys offered as an explanation for these phenomena that the living leukocytes in the original exudate secreted alexin or complement which enhanced the bactericidal activity of the exudate, that the leukocytes, subsequently added to 9 Buchner. Munch, med. Woch., No. 24, 1894. 10 Hahn. Archiv f. Hyg., Vol. 25, 1895. 11 Denys and Kaisin, Denys and Havet. La Cellule, Vol. 9, 1893; Vol. 10, 1894. RELATION OF LEUKOCYTES TO IMMUNITY 301 inactivated exudate, however, had lost vitality during the process of isolation and washing, and no longer possessed secretory power. Hankin,12 Kanthack and Hardy 13 had gone even farther than this, and had attributed the production of alexin to the eosinophile leukocytes particularly. Metchnikoff,14 basing his opinion on his own studies, those of his pupils, and many other investigations similar to those mentioned above, came to the conclusion that, under ordinary conditions, the normal blood contains no free bactericidal substances. He assumes that these substances are entirely intracellular, being constituents of the various phagocytic elements, by means of which the cells digest the foreign elements they take up. He believes that there are essen- tially two varieties of such digestive enzymes or "cytases" — just as there are two varieties of phagocytes. The microphages, chiefly con- cerned in the digestion of bacteria, secrete the bactericidal alexin, or, as Metchnikoff calls it, "microcytase." The macrophages, the large mononuclear lymph and endothelial cells, primarily concerned in the phagocytosis of cellular elements (red cells, etc.), contain another variety of digestive enzyme, the "macrocytase," or cytolytic (hemo- lytic) alexin. The supposition that the hemolytic "cytase" is largely derived from the macrophages was based particularly upon the in- vestigations of Metchnikoff's pupil, Tarassewitch,15 who found that the extracts obtained from lymph nodes, and other organs rich in macrophages, possessed hemolytic properties. Both this work and the preceding studies regarding the extraction of alexin from poly- nuclear leukocytes will be more fully discussed below. Maintaining that these cytases are purely intracellular under ordinary conditions, Metchnikoff believes that, in normal animals, the destruction of invading bacteria or of injected cellular substances (blood cells, etc.) is accomplished entirely by the phagocytic process, with subsequent intracellular digestion. In immunized animals, however, there is present in the circulating blood another substance, not identical with the cytases, but also derived from the leukocytes or from the blood-forming organs — the "fixateur" (Ehrlich's "ambo- ceptor" — Bordet's "sensitizer"). This specific "fixateur" sensitizes the bacteria or other antigens to the action of the cytases. For his assumption regarding the origin of this sensitizer he finds support largely in the researches of Pfeiffer and Marx, and others mentioned in our section on the origin of antibodies, as well as in the simi- lar investigations of Deutsch,16 carried on under Metchnikoff ?s per- sonal supervision. 12Hankin. Centralbl. f. Bakt., Vol. 12, 1892. 13 Kanthack and Hardy. Proc. Eoy. Soc., Vol. 52, 1892. 14 Metchnikoff. Ann. de I'Inst. Pasteur, Vol. 7, 1893; Vol. 8, 1894; Vol. 9, 1895. 15 Tarassewitch. Ann. de I'Inst. Past., Vol. 16, 1902. 16 Deutsch. Ann. de I'Inst. Pasteur, Vol. 13, 1899. 302 INFECTION AND RESISTANCE Final digestion of the sensitized antigens (bacteria or blood cells), however, can take place only under the influence of the cytases intracellularly, unless by previous leukocytic injury these enzymes have been liberated into the blood stream. While it is admitted, then, that bacteria may be killed and di- gested both intra- and extracellularly in the animal body, the cytases, which accomplish this, are regarded as the same in both cases, being derived from the phagocytic cells. In immunized animals "fixateur" may be produced under the stress of active immunization and fur- nished to the blood stream by the blood-forming organs. By this substance bacteria and cells may be sensitized. However, the enzyme by which digestion is actually accomplished, "cytase" or alexin, is not present free in the blood even in immune animals unless it has become free and extracellular by injury to the leukocytes. How, then, on this basis does Metchnikoff account for the Pfeiffer phenomenon, in which the extracellular destruction of bacteria takes place rapidly in the peritoneal exudate ? His explanation is the fol- lowing: When bacteria or other substances are injected into the peritoneum there is a preliminary injury of leukocytes (phagolysis), and by this alexin or cytase is liberated. When cholera spirilla, for instance, are injected into the peritoneal cavity of an immunized guinea pig there follows a short period during which few if any leukocytes are present in the exudate, but many may be found gath- ered in motionless clumps in the folds of the peritoneum and mesen- tery, incapable of phagocytosis and apparently injured. If such leukocytic injury can be avoided Metchnikoff claims that the extra- cellular lysis of cholera spirilla will fail to take place. Thus if sterile broth or salt solution is injected into the peritoneum of a guinea pig a preliminary phagolysis will be followed by an accumu- lation of leukocytes. If cholera spirilla are now introduced no extracellular digestion is seen, but, instead of this, rapid phagocytosis takes place. This he says is due to the fact that the freshly accumu- lated, healthy phagocytes, collected in response to the preliminary broth injection, are not easily injured and do not undergo phago- lysis; no cytase is liberated and, in consequence, no serum bacterio- lysis can take place. In the same way he claims that the hemolysis of red blood cells (goose blood) in the peritoneum of specifically immunized guinea pigs may be prevented if, by a previous injection of broth, healthy leukocytes have been caused to accumulate. In such a case the goose blood corpuscles are rapidly ingested by the phagocytes and no hemolysis occurs. It is self-evident that the validity of this interpretation of the occurrences is strictly dependent upon the demonstration that the circulating blood normally contains no alexin or complement. This is rigidly maintained by Metchnikoff, and is indeed one of the most important uncertainties of serology. He asserts that alexin appears RELATION OF LEUKOCYTES TO IMMUNITY 303 in the blood serum only because changes in the leukocytes take place during coagulation. It is not, by any means, settled that Metchni- koff is right in this — in fact, more recent investigations seem to show that he is wrong, and that we may assume definitely that alexin is present in considerable amounts in the circulating blood plasma of normal animals. Metchnikoff s denial of this is based chiefly on the experiments of Gengou. Gengou 17 took the blood from various animals into paraffined tubes and centrifugalized it at low temperature before it could clot. This freed the plasma from the cells before clotting, though coagulation of course took place as soon as this plasma was removed to tubes and kept at room temperature. The serum ex- pressed from this clotted plasma he compared for alexin contents (bactericidal properties) with that obtained from clotted whole blood. He found that, in all cases examined (dogs, rabbits, and rats), the plasma contained practically no bactericidal substances, or at any rate an incomparably smaller amount than was present in the serum obtained from the clotted blood. These experiments of Gengou would be conclusive in establishing Metchnikoff's theory if they were confirmed by other observers. This, however, has not been the case. Fetter son 18 found no differ- ence between the bactericidal properties of serum and oxalate plasma, and Lambotte 19 arrived at similar results when he compared serum with the coagulable plasma obtained by tying off a section of a vein and centrifugalizing the blood without opening the vessel. Hew- lett,20 Falloise,21 Schneider,22 and more recently Dick 23 and Addis,24 whose work has been done with particular attention to technical accuracy, fail to confirm Gengou, finding no appreciable difference between plasma and serum. In favor of Gengou's results are the investigations of Herman 25 and the more recent ones of Gurd.26 Further supporting Gengou's conclusion is the observation recorded by a number of workers (Wal- ker,27 Longcope,28 and others) that the complement or alexin con- tents of serum will increase somewhat as the serum is allowed to 17 Gengou. Ann. de I'Inst. Past., Vol. 15, 1901. 18 Petterson. Arch. f. Hyg., Vol. 43, 1902. 19 Lambotte. Centralbl. f. Bakt., Vol. 34, 1903. 20 Hewlett. Zeitschr. f. Heilkunde, 24, 1903. 21 Falloise. Bull, de I'Acad. Eoy. de Med., 1905. 22 Schneider. Arch. f. Hyg., 65, 1908. 23 Dick. Jour. Inf. Dis., Vol. 12, 1913. ** Addis. Jour. Inf. Dis., Vol. 10, 1912. 25 Herman. Bull, de I'Acad. Roy. de Med., 1904. 26 Gurd. Jour. Inf. Dis., Vol. 11, 1912. 27 Walker. Jour, of Hyg., 3, 1903. 28 Longcope. Med. Bull. Univ. of Pa., 1902, Vol. 15, p. 331. 304 INFECTION AND RESISTANCE stand on the clot. This observation, too, has been rendered incon- clusive by contrary reports from other investigators. Longcope,29 further, has found that alexin was more plentiful in the blood of individuals suffering from leukemia — in which of course a larger percentage of leukocytes is present in the circulation. This, too, has been contradicted by other workers, but even if upheld would not influence the possibility of there being alexin in the normal circula- tion. On the whole Gengou's contentions with their consequent bearing upon MetchnikofFs theory cannot be accepted as final. In fact, the greater part of available experimental evidence seems to point to the actual presence of alexin in the normal circulating blood. This seems also indicated by the unquestionable fact that active phagocytosis may take place in the circulation of an animal and, as we shall see below, free alexin is probably necessary (as opsonin) in this process. Further evidence in this direction also is furnished by the immediate anaphylactic shock which follows the injection of antigen into the blood stream of a sensitized animal, a process in which we have much reason to believe that alexin takes an active part. However, the problem is a difficult one, and, while we favor the opinion that free alexin is present in the intravascular blood, we must admit that a crucial experiment has not yet been formulated. Now, as regards the apparent extraction of alexin from leuko- cytes and lymphatic organs, we have already called attention to the fact that most of the researches associating these cells with the bac- tericidal substances were carried out before the dual mechanism of sensitizer and alexin in bacteriolysis had been fully worked out. In consequence conclusions were formulated from the mere facts of the presence of bactericidal or hemolytic properties in cell-extracts with- out the further determination of heat stability or the possibility of reactivation. Most of the earlier work also was done without suffi- cient attention to the separation of the cells and the serum of leuko- cytic exudates. The first one to do this carefully was Hahn,30 who, like his predecessors, concluded that the bactericidal leukocytic sub- stances, undoubtedly encountered by him, were identical with alexin. Doubt was first cast upon this by Schattenfroh,31 who worked with leukocytes suspended and extracted in physiological salt solution. He found that bactericidal substances were, indeed, obtained in this way, but that, unlike alexin, these substances were thermostable, withstanding exposure to a temperature of 56° C. and destroyed only by exposure to temperatures as high as 75° 0. to 80° C. for thirty minutes. 29 Longcope. Jour, of Hyg.} Vol. 3. 30 Hahn. Arch. f. Hyg., Vol. 25. 31 Schattenfroh. Arch. f. Hyg., Vols. 31 and 35, 1897. RELATION OF LEUKOCYTES TO IMMUNITY 305 Moxter,32 a little later, working with cholera spirilla, also came to the conclusion that the leukocytic bactericidal substances were not identical with those found in the blood serum. Fetter son,33 too, made thorough investigations into the nature of the bactericidal substances extracted from the leukocytes, and, working chiefly with B. proteus and B. anthracis, found such substances in the leukocytes of dogs, rabbits, and guinea pigs active against the bacteria mentioned above, but failed to find them active, at least in guinea pig and cat leuko- cytes, against B. typhosus or the cholera spirillum. He expresses the opinion that bactericidal leukocytic substances are normally given up to the blood in minute quantity only or not at all, and that these substances hold no definite relationship to the bactericidal sub- stances found in blood serum. In a later investigation he showed that the "endolysins," as he now calls the leukocytic bactericidal substances, may, like many enzymes and serum bacteriolysins, be precipitated out of solution with alcohol and ether ; but he separates them absolutely from serum lysins and complement. The latter, while they may be, in part at least, secreted by the leukocytes, are, according to Pe'tterson, induced easily to come out of the cells during life by slight injury or other stimulation, while the endolysins them- selves are abstracted from the cells only after extensive extraction or maceration. Schneider 34 has come to similar conclusions and speaks of the endocellular bactericidal substances as "Leukine." In a recent in- vestigation of the same subject the writer 35 has in all essentials con- firmed Schattenf roh's original conclusions regarding the heat sta- bility of the extracted leukocytic bactericidal substances, and has shown that after inactivation by heat these substances are not reacti- vable by the addition of fresh leukocytic extracts, and that the yield obtained from the leukocytes of immunized animals is not greater than that obtained from normal leukocytes. It appears, therefore, that, contrary to Metchnikoff's first sup- position, the enzymes which bring about the digestion of phagocyted bacteria within the cell are not identical with those which bring about a similar extracellular digestion. In addition to the demon- stration of a definitely different structure possessed by the bacteri- cidal leukocytic extracts, as evidenced by their heat stability, we have the negative evidence that neither true alexin nor sensitizers have ever been successfully extracted from such cells. It is still possible that this may eventually be done — but, al- though indirect evidence like that of Denys, Longcope's observations in leukemia, and the occasional increase of the alexic, powers q£ serum 32 Moxter. Deutsche med. Woch., 1899, p. 687. 33 Petterson. Centralbl f. Bakt., i, 39, 1905; 46, 1908. 34 Schneider. Archiv f. Hyg., Vol. 70, 1909. 35 Zinsser. Jour. Med. Res., Vol. 22, 1910. 306 INFECTION AND RESISTANCE after standing on the clot points to a probability of this, no direct evidence has so far been satisfactorily produced. In the hope that the leukocytes would give up alexin — possibly as a secretion as sug- gested by Denys — the writer, with Cary, some years ago kept washed leukocytes at 37.5° C. in Locke's solution, but was unable to find any evidence of alexin production within 48 hours. The apparent extraction of hemolysin from macrophages by Tarassewith, moreover, has met with a similar refutation. Korschun and Morgenroth 36 have shown that these hemolytic extracts are ex- tremely heat resistant, are alcohol and ether soluble, and do not act as antigens. They are quite different from the serum hemolysins, therefore, and probably closely related to the hemolytic lipoidal substances described by Noguchi and others. Further strong arguments against the assumption of the pres- ence of hemolytic alexin in the body of the macrophages have been advanced by Gruber 37 and by Neufeld.38 Gruber showed that no extracellular hemolysis takes place when leukocytes are brought together with sensitized red blood cells, and Neufeld showed that even after the phagocytosis of such sensitized cells the hemolysis is very much slower, and of a different character from extracellular serum hemolysis. In the intracellular digestion there are no mere solution of the hemoglobin and formation of a shadow form (stroma), but there occur a gradual degeneration with the forma- tion of a granular detritus of hemoglobin. It is probable, then, that the digestion of bacteria and cells within the phagocytes is carried out by substances not identical with those taking part in serum lysis. It is not unlikely that the intra- cellular process is a quite complicated one, not depending on a single enzyme. In addition to the bactericidal substances extracted from leuko- cytes a number of true enzymes have indeed been obtained by various investigators. We have mentioned in another place that one of the earliest observations in this respect was that of Leber,39 who noticed that sterile pus could liquefy gelatin. It may be commonly observed in paraffin or celloidin sections of staphylococcus abscesses that a ring of apparently digested or degenerating tissue is formed about an accumulation of leukocytes — in foci in which the bacteria may be too sparse to be held accountable for the changes. These leuko- proteases have subsequently been carefully studied by Opie,40 Joch- mann and Miiller,41 and a number of others. 36 Korschun and Morgenroth. Berl. klin. Woch., 1902. 37 Gruber. Quoted from Sachs, in "Kraus u. Levaditi Handbuch," Vol. 2,. p. 991. 38 Neufeld. Arb. a. d. kais. Gesundh. Amt., Vol. 28, 1908. 39 Leber. "Die Entstehung der Entziindung," Leipzig, 1891. 400pie. Jour. Exp. Med., Vol. 7, 1905; Vol. 8, 1906; Vol. 9, 1907. 41 Miiller and Jochmann. Munch, med. Woch., Nos. 29 and 31, 1906. RELATION OF LEUKOCYTES TO IMMUNITY 307 Opie found that two distinct proteolytic enzymes could be ex- tracted from the cells of exudates obtained by turpentine injections. One — peculiar to the polynuclear leukocyte, and similar to one pre- viously described by Miiller 42 — acts in a weakly alkaline medium. The other, present particularly in exudates containing a predominat- ing number of mononuclear cells, acts in a weakly acid reaction. Tschernorutski also found protcolytic ferments in both micro- and macrophages, but found no lipase in the polynuclear extracts. This seems particularly interesting in view of the great resistance to intracellular digestion noticed in acid-fast bacteria, a point of some importance in connection with the destruction in the body of such micro-organisms as the bacilli of tuberculosis and leprosy.43 Joch- mann 44 states that the action of the leukoprotease, which acts in an alkaline medium upon casein, results in the formation of tryptophan and ammonia, and believes it to be functionally very similar to trypsin. It is interesting to note that the most active protease is obtained from pus as it forms about acute infections or other stimuli which lead to the accumulation of polynuclear leukocytes, whereas it is apparently completely absent from tuberculous pus. The question immediately arises, are these leukoproteases identi- cal with the bactericidal substances extracted from leukocytes as de- scribed above ? For it might well be that bacterial death resulted merely from the digestive action of the enzyme upon the bacterial protein. Jochmann,45 who has approached this problem experi- mentally, has answered it in the negative. By repeated alcohol precipitation of glycerin extracts of leukocytes he obtained an en- zyme preparation which possessed absolutely no bactericidal prop- erties, though it was still actively proteolytic. Not only did this relatively pure ferment possess no bactericidal action, but bacteria actively proliferated when suspended in it. Joch- mann believes that living bacteria are not amenable to the enzyme possibly because of their possession of an "antiferment," at least this would follow in some cases from the experiments of Kantoro- wicz.46 The leukoproteases, therefore, appear to possess no direct sig- nificance in bacterial immunity. Their function seems rather to lie in the resorption of dead tissues, fibrin, blood clots, etc. Friedrich Miiller 47 has pointed out their possible importance in the rapid de- struction and liquefaction of the massive fibrinous exudates remain- ing after the crisis in lobar pneumonia. 42 Miiller. Congr. f. inn. Med., Wiesbaden, 1902. 43 Zinsser and Gary. Jour. A. M. A., 1911. 44 Jochmann. Leucozyten Fermente, etc., "Kolle u. Wassermann Hand- buch," 2d Ed., Vol. 49, 2. 45 Jochmann. Zeitschr. f. Hyg., 61, 1908. 46 Kantorowicz. Munch, med. Woch., No. 28, 1909. 47 Friedrich Miiller. "Verhand. d. Kongr. f . inn. Med.," 1902. 308 INFECTION AND RESISTANCE From the discovery of antibacterial properties in the extracts of leukocytes it is but a logical step to the attempt to utilize these sub- stances therapeutically. This was especially called for in view of the disappointing results which have attended the injection of even large amounts of bactericidal sera into animals and human beings in whom anthrax bacilli, streptococci, or any other of the invasive bac- teria or true parasites had gained a foothold. Petterson 48 was probably the first to study this phase of the problem systematically in connection with anthrax infection in dogs and rabbits. In pre- liminary studies he claimed to have determined that when leukocytes are left in contact with serum for four hours or longer there develops in the mixture a bactericidal power far superior to that which is possessed by these elements when separately kept in salt solution and mixed only just before the bactericidal tests. He attributes this to the fact that in dogs, at least, the leukocytes furnish bactericidal substances to the serum — an assumption which is entirely in accord with the earlier opinion of Denys and Kaisin,49 which we have men- tioned in another place. In direct continuance of these experiments he injected leukocytes into dogs at the same time at which he in- fected them with anthrax and observed a moderately protective in- fluence, which, however, he admits was not very great. He followed this work in 1906 with similar observations on the protective influ- ence of leukocytes in intraperitoneal infections of guinea pigs with typhoid bacilli. In these experiments 50 he made the curious ob- servation that, although such protective influence was unquestionable, the guinea pig leukocytes contained no bactericidal substances active against typhoid bacilli. In consequence he concluded that the de- struction of these bacteria in the guinea pig was due entirely to the serum-antibodies absorbed by the micro-organisms before phago- cytosis, even when the actual destructive process took place intra- cellularly. The protective effect following on the injection of the leukocytes he attributed to an indirect influence of the leukocytic substances in stimulating the more rapid accumulation of alexin or complement in the peritoneum, with consequently more powerful phagocytosis. Following this, in 1908, Opie51 carried out experi- ments in which he observed that leukocytes injected intrapleurally into dogs, together with tubercle bacilli, exerted a distinct protection in that the course of the disease was prolonged and the tendency toward healing was more pronounced than in the controls. In the same year extensive observations on the protective prop- erties of leukocyte extracts were published by Hiss. 48 Petterson. Centralbl. f. Bakt., Vol. 36. 1904. 49 Denys and Kaisin. "La Cellule," Vol.' 9, 1893. 50 Petterson. Centralbl f. Bakt.., Vols. 40 and 42, 1906. 51 Opie. Jour. Exp. Med., 1908. RELATION OF LEUKOCYTES TO IMMUNITY 309 Hiss 52 worked at first with extracts of dog, rabbit, and guinea pig leukocytes; later he confined himself entirely to rabbit leuko- cytes. He extracted the leukocytes at first by repeated freezing and thawing in physiological salt solution, but the technique of his sub- sequent work was uniformly as follows: Intrapleural injections of aleuronat emulsions were made in rabbits and, after about 24 hours, the resulting exudates were taken away with sterile pipettes and centrifugalized before clotting could take place; the serum was de- canted and the leukocytes then emulsified in distilled water, in quan- tity about equal to the amount of serum poured off. In this the leuk- ocytes were allowed to stand for a few hours at incubator tempera- ture, and then in the ice-box until used. For his experimental work' in both animals and man, in most instances, not only the clear super- natant fluid was injected, but the cell residue as well; since Hiss realized that the extractions were necessarily incomplete. In in- travenous work, of course, the supernatant fluid alone was injected. With leukocytic extracts so prepared Hiss treated staphylococ- cus, typhoid bacillus, pneumococcus, streptococcus, and meningococ- cus infections in rabbits and obtained results which justified him in concluding that the leukocyte extract exerted strong protective action in all of these cases. Many of his animals survived infections fatal to controls even when the treatment was delayed as long as 24 hours after infection. Subsequently Hiss and Zinsser 53 treated series of patients, ill with pneumonia, meningitis, and staphylococcus infec- tions, with leukocyte extracts prepared by the method of Hiss, and felt that they were justified in concluding that in many cases, at least, the course of the disease was favorably influenced by the leuko- cyte extract. Favorable results have since then been obtained also by Lambert in erysipelas, and by Hiss and Dwyer in a variety of conditions. Dwyer has used the extract in various infections of the eye, ear, nose, and throat. While there seems to be little question about the actually favor- able influence of the leukocyte extract, both in experiments with animals and in the treatment of human cases, there has been consid- erable difficulty in determining the reasons for this influence. In subsequent studies Hiss and Zinsser (loc. cit.) were able to show that the extracts did not favor phagocytosis and that the moderate bacteri- cidal properties possessed by the leukocytic substances could not ac- count for their effectiveness. There did seem to be a more rapid accumulation of phagocytes in the peritoneal cavities of guinea pigs infected with cholera spirilla when leukocyte extract was injected with the bacteria, and it is not impossible, in fact, it seems probable to the writer, from subsequent experience, that the protective prop- 52 Hiss. Jour, of Med. Res., new series, Vol. 14, 1908. 53 Hiss and Zinsser. Jour, of Med. Ees., new series, Vol. 14, 1908. 310 INFECTION AND RESISTANCE erties of the leukocyte extracts are attributable, in part at least, to their positively chemotactic effect. The entire problem opened up by the work of Hiss cannot be regarded as settled. Observations, both experimental and clinical, are still in progress, and it is hoped that the next few years may definitely decide in how far this treatment is applicable to human cases. It is not easy to draw conclusions from clinical observations since it is impossible to parallel such cases with untreated controls; in consequence the truth can be elucidated only by a multiplication of statistics. While the writer and others have treated a great many cases with disappointment, again the striking results occasionally observed have been so encouraging that it seems of the utmost im- portance to give the treatment extensive trial, especially since many injections have been made without any harm whatever to the patients. Although the experience thus far gathered permits of no definite conclusions, the writer would suggest from his own experience and his observation of that of others that the use of the leukocyte extract of Hiss be confined for the present to diseases like erysipelas, menin- gitis, and the pyogenic infections in which the process is distinctly localized and no general septicemia has supervened. It should also be given a thorough trial in broncho- and lobar pneumonia in which the bacteriemia which occurs represents very probably a constant dis- charge into the blood stream of bacteria from the pneumonic focus, rather than the firm establishment of bacterial growth within the blood itself. With few exceptions absolutely no results seem to have followed its use when such a septicemia has become established. In the class of cases first mentioned, however, where a localized infec- tion has been obst'inate and unusually violent, many brilliant results have been obtained. Judging from the results of Dr. Adrian Lam- bert, and more recent ones obtained by Dr. Dwyer, we would have no hesitation in stating that erysipelas is favorably influenced in most cases. The above suggestions are made since it seems that in the question of clinical therapy much delay in the proper estimation of the value of a new type of treatment can be avoided by an intelligent choice of cases. CHAPTER XIII FACTORS DETERMINING PHAGOCYTOSIS OPSONINS, TKOPINS FROM the very beginnings of his researches upon phagocytosis Metchnikoff recognized that the process was profoundly influenced by the properties of the fluid constituents of the blood plasma in which the phenomenon occurred. Both he and his pupil Bordet,1 at this time working in Metchnikoff's laboratory, noticed that the phagocytic activity of leukocytes was greater in immune than in normal sera and associated this with the specific properties of the immune substances or antibodies in these sera ; Metchnikoff himself interpreted the phagocytosis-enhancing power of the serum as a stimulation of the leukocytes and referred to the serum constituents by which this effect was produced as "stimulins." A closer analysis of the factors involved in this interrelationship, however, was not attempted at this time by him or his pupils, although indirect ref- erence was made to it in a number of articles emanating from this school in the course of investigations on kindred problems of phago- cytosis. Thus Gabritschewsky,2 in 1894, published a paper on "Leukocytose dans la Diphtheric/' in which he concluded that the poison of diphtheria bacilli, among other harmful effects, diminished the phagocytic power of the leukocytes, and that one of the beneficial influences of the curative serum was to render these and other cells "less sensitive to the bacterial poisons.7' This may be interpreted as indicating an assumption that the action of an immune serum in increasing phagocytic activity rested rather upon its influence upon the bacterial products than upon any stimulation of the phagocytes themselves. However, in diphtheria the action of the leukocytes was, even at this time, recognized as a merely secondary one, and Gabritschewsky's results did not materially influence the "stimulin" conception. The first extensive investigation which occupied itself directly with these problems was that of the Belgian bacteriologists Denys and Leclef.3 The publication of these workers deals primarily with 1Bordet. Ann. de Vlnst. Past., 1895. 2 Gabritschewsky. Ann. de I'Inst. Past.., 1894. 3 Denys and Leclef. La Cellule, 11, 1895. 311 312 INFECTION AND RESISTANCE the nature of streptococcus immunity in rabbits. It established, first of all, the paramount importance of phagocytosis in the resistance of animals against these bacteria, and made clear that the destruction of bacteria was carried out equally as well by the leukocytes of nor- mal as by those of immune animals, but was powerfully enhanced when either normal or "immune" leukocytes were combined with immune serum. Their work, therefore, indicated again that the in- creased phagocytosis of virulent bacteria, taking place in immune animals, depended clearly upon alterations in the functions of the serum rather than in those of the cells, and they suggested that the influence of this serum was not necessarily one of leukocyte stimulation, but might rather consist in action upon the bacteria, rendering them less resistant to phagocytosis. They say in sub- stance: "A notre avis, on pourrait tout aussi bien admettre que la substance vaccinante ou antitoxique agit, non pas sur le leukocyte, mais sur un poison renferme dans le corps du microbe ou dissous dans le milieu, et qui preserve le micro-organisme centre les at- teintes du leukocyte." 4 In this statement we have, in brief, the distinct formulation of our present view of the "opsonins." 5 Observations with pneumococci and streptococci carried out after this by Marchand 6 and by Mennes,7 whose investigations we cannot discuss in detail, beside confirming most of the observations of Denys and Leclef, brought out especially the relation of the virulence of micro-organisms to phagocytosis, showing that very virulent strains were taken up to a slight degree only in the presence of normal serum, but were subject to active phagocytosis when im- mune serum was employed. This, too, seemed to point primarily to the fact that the serum influenced rather the bacteria than the phagocytes, although no convincing proof is brought for this in their publications. Though much that had bearing indirectly on thia problem was written during the following years, no definite progress was made beyond the results of Denys and his pupils until 1902, 4 In our opinion one can just as well believe that the vaccinating or anti- toxic substance acts not upon the leukocyte but upon a poison inclosed within the body of the bacteria or dissolved in the medium, which preserves the micro-organism against the attacks of the leukocyte. 5 Denys formulated this view with still greater clearness and positiveness at the Congress of Hygiene held at Brussels in 1903. We take our citation from the discussion on opsonins by Gruber (3d meeting Freie Vereinigung f. Mikrobiol., Vienna, 1909, Centralbl f. Bakt., I Ref., Vol. 44, Suppl. p. 3). Following is Denys' statement: 1. The phagocytosis in immune sera is de- pendent upon substances which are precipitated with the euglobulins. 2. These substances cause phagocytosis by inciting a physical alteration of the micro-organisms. 3. These substances are specific. 6 Marchand. Arch, de Med. Exp., 1898. 7 Mennes. Zeitschr. f. Hyg., Vol. 25. FACTORS DETERMINING PHAGOCYTOSIS 313 when Leischman 8 introduced a technique by means of which it be- came possible to observe the process of phagocytosis with fresh serum and leukocytes in vitro. By utilizing this technique and improving upon it Wright and Douglas in the following year (1903) evolved a method by means of which phagocytic activity could be quantitatively measured with reasonable accuracy. They worked at first with staphylococcus phagocytosis by human leukocytes in the presence of human citrate plasma, a research undertaken primarily because Wright,9 in collabo- ration with Windsor, had previously determined that human blood serum possessed practically no bactericidal power for this organism, and that phagocytosis was probably the chief mechanism of protection which the human body possessed against these bacteria. The re- searches of Wright and Douglas 10 were carried out chiefly by mixing equal volumes of bacteria, serum, and leukocytes (in citrate sus- pension),11 allowing these elements to remain together at 37.5° C. for varying periods, then staining on slides and determining the degree of phagocytosis by counting the numbers of bacteria taken up by each polynuclear leukocyte. Though many technical difficulties had to be overcome, and although the method at its best still permits of much personal error, careful work and untiring repetition made possible a considerable degree of accuracy, and definite facts regard- ing the mechanism of • phagocytosis, heretofore merely suspected, could be recorded. The most important result of these investiga- tions was the unquestionable establishment of the function of serum in the process of phagocytosis, namely, that it in no way "stimu- lated" the leukocytes in the sense of Metchnikoff, but rather acted entirely upon the bacteria, preparing them for ingestion. For this reason Wright coined the word "opsonins" (6^oi>«o= I prepare food) for the serum constituents which brought about this effect, believing them to be new antibodies, entirely distinct from the other serum antibodies heretofore recognized. Wright and his followers now concluded that the role of the leukocyte in taking up bacteria was entirely dependent upon the opsonin contents of the serum. In a menstruum containing no serum, or in a serum in which the opsonins had been destroyed by heat, they found practically no phagocytic action on the part of washed serum-free leukocytes, and they, therefore, doubted the oc- currence of spontaneous phagocytosis on the part of leukocytes themselves. 8 Leischman. Brit. Med. Jour., Vol. 2, 1901, and Vol. 1, 1902. 9 Wrig-ht and Windsor. Jour, of Hyy., Vol. 2, 1902. 10 Wright and Douglas. Proc. Roy. Soc., 72, 1903, 73 and 74, 1904. See also Wright, "Studien fiber Immunisierung," Fischer, Jena, 1909. 11 At first bacteria were merely mixed in equal volumes with citrated whole blood. 314 INFECTION AND RESISTANCE In this point it is not unlikely that Wright is mistaken, since other observers, notably Lohlein,12 have observed the phagocytosis of various bacteria by washed leukocytes in indifferent, opsonin-free media. Although we may take it as assured that such spontaneous phagocytosis may take place (Metchnikoff and a number of others having obtained results similar to those of Lohlein), this is prob- ably never very intense. In fact, Wright, in some of his later work, does not insist rigidly upon the non-occurrence of spontaneous phagocytosis, but attempts to associate such phenomena with the salt contents of the medium in which it occurs. Together with Reid,13 he determined that spon- taneous phagocytosis of tubercle bacilli unquestionably takes place, is most intense at a concentration of about 0.6 per cent. NaCl, and diminishes as the concentration is increased. This, as we shall see, has bearing on the possible physical explanations advanced to ac- count for opsonic action, and has its parallels in experiments on the (influence of electrolytes on agglutination and precipitation. The fact remains that Wright demonstrated by his work that MetchnikofFs original view, which interpreted the difference be- tween susceptibility and immunity as a difference between the in- herent phagocytic powers of the leukocytes, is incorrect, and that the essential regulating influence affecting phagocytosis rests upon the action of the serum upon the bacteria. The following experiment from the work of Hektoen and Rue- diger 14 illustrates this point with exceptional clearness. It shows that human leukocytes in the presence of normal defibrinated blood will take up bacteria energetically. When the leukocytes, however, are washed free of blood and added to untreated bacteria phago- cytosis is practically nil. If, however, such washed leukocytes are mixed with bacteria that have been previously in contact with serum active phagocytosis will take place. In other words, the bacteria have been altered by the serum in such a way that they are now amenable to phagocytosis by washed leukocytes. The serum then acts upon the bacteria and not upon the leukocytes. TABLE II Phagocytosis by Human Leukocytes of Sensitized Bacteria Average Phagocytosis Human leukocytes (defibrinated blood) + Staphylococcus aureus 22 . Human leukocytes (washed in NaCl solution) + Staphylococcus aureus. 1 . 2 Human leukocytes (washed in NaCl solution) -f Staphylococcus aureus (treated with human serum) 10 . Human leukocytes (defibrinated blood) -f Streptococcus 300 22 . 12 Lohlein. CentralbL f. Bakt., 38, 1906, Beihef t, p. 32 ; also Munch, med. Woch., 1907, p. 1473. is Wright and Reid. Proc. of Royal Soc. B., Vol. 77, 1906. 14 Hektoen and Ruediger. Jour. Inf. Dis., Vol. 2, 1905, p. 132. FACTORS DETERMINING PHAGOCYTOSIS 315 Average Phagocytosis Human leukocytes (washed in NaCl solution) -f- Streptococcus 300 .... 1 . Human leukocytes (washed iu NaCl solution) -f Streptococcus (treated with human serum) 14 . Human leukocytes (washed in NaCl solution) + Streptococcus (treated with guinea pig serum) 12 . Human leukocytes (washed in NaCl solution) + Streptococcus (treated with rabbit serum) 14. Wright and Douglas' 15 work was done at first with normal serum or normal citrate plasma, and in this case they found that the opsonins were essentially unstable, being easily weakened by ex- posure to light, or heat, and even when preserved in sealed tubes in the dark they diminished noticeably on standing for 5 or 6 days. Other writers who have worked with the opsonic substances in nor- mal serum have confirmed this instability of the normal opsonin, although even Wright himself admits that heating to 60° C. does not entirely destroy the opsonic power, though it reduces it to a minimum. A protocol from Wright and Douglas' first paper will best illustrate the degree of reduction of opsonic power resulting from the exposure of normal serum to 60-65° C. for 10 to 15 minutes. A. Unheated serum Wright — Staphylococcus suspension 1 vol. — Blood cells Wright 3 vols. (1) Phagocytic average 20 cells 17 .4 (2) Phagocytic average 20 cells 19 .8 B. Heated serum as above. (1) Phagocytic average 52 cells 0.6 (2) Phagocytic average 46 cells 3.4 The experiments just cited refer only to the opsonic powers of normal serum. When an animal is immunized with any particular micro-organism or other cellular antigen, such as red blood cells, etc., a marked specific increase of opsonins occurs, but unlike the opsonins of normal serum these newly formed elements in the im- mune serum seem to possess a much greater resistance to heat. Neufeld and Rimpau,16 who have studied these constituents of immune serum with especial thoroughness, have shown that heating to 62° to 63° C. for as long as three-quarters of an hour does not destroy them, and that such sera may be preserved for as long as several years without their complete disappearance.17 We may accept as definitely determined, therefore, that there is a qualitative difference between the serum components which initiate phagocytosis in normal serum (normal opsonins) and those which carry out the same function to a much more powerful degree in 15 Wright and Douglas. Cited in Wright, "Studien iiber Immun., etc.," p. 9. 16Neufeld and Rimpau. Deutsche med. Woch., No. 40, 1904; Zeitschr. f. Hyg., Vol. 51, 1905. 17 Leishman. Trans. London Path. Soc.} Vol. 56, 1905. 316 INFECTION AND RESISTANCE immune serum. This is the more surprising since, in the case of all other antibodies (lysins, agglutiiiins, etc.), it has been shown that in structure and mode of action the antibodies of immune serum are in every way qualitatively similar to the corresponding ones of nor- mal serum,18, 19 representing merely a specific quantitative increase of substances originally present in small amount. This difference between the normal and immune opsonic sub- stances has added much difficulty to the investigation of the nature of these bodies, and we may approach the problem with greater clearness by considering them separately, at first, attempting to define the relations between them after we have set down the facts ascertained in connection with each. In their earliest investigations upon the normal opsonins Wright and Douglas 20 regarded them as new antibodies, separate and dis- tinct from those already known. There is no convincing proof of this, and a number of other interpretations of the observed phe- nomena are possible. Indeed, the burden of proof is rather upon those who would establish the existence of a new antibody, for before this can be done it must be shown that the new function is not merely another property of the serum constituents already known. For, as Gruber has justly said, "One of the most important attributes of the natural scientist is economy of hypotheses." And in the case of the normal opsonins there are many good reasons for regarding them as possibly identical with known serum constituents. The two possi- bilities suggested have been (1) Are the opsonic substances identical with the alexin or complement? or (2) Do they represent the com- bined action of the normal sensitizer of the serum activated by the alexin ? The similarity of normal opsonin with alexin or complement has been brought out especially by Muir and Martin,21 by Baecher,22 and by Levaditi and Inmann.23 The fact that both are thermolabile has been mentioned above. In addition to this, as Muir and Martin24 have shown, all antigen- antibody complexes which absorb alexin out of serum at the same time remove the normal opsonin. Thus sensitized red corpuscles, sensitized bacteria, and specific precipitates added to normal serum take out its opsonic substances. From this fact they also concluded that the normal opsonins like alexin were non-specific. For just as 18 Dean. Proc. Royal Soc., 76, 1905. 19 Neuf eld and Hiine. Arb. a. d. kais. Gesundh. Amt., Vol. 25, 1907. 20 Wright and Douglas. Loc. cit. 21 Muir and Martin. Br. Med. Jour., Vol. 2, 1906; Proc. Royal Soc. B., Vol. 79, 1907. 22 Baecher. Zeitschr. f. Hyg., Vol. 56, 1907. 23 Levaditi and Inmann. C. R. de la Soc. BioL, 1907, pp. 683, 725, 817, 869. 24 Muir and Martin. Loc. cit. FACTORS DETERMINING PHAGOCYTOSIS 317 the alexin of a serum may serve to activate a considerable variety of sensitized antigens, so the opsonic action of a normal serum may functionate upon a large variety of bacteria. Muir and Martin were probably wrong in this and, as we shall see below, normal opsonins, like normal sensitizers, may be regarded as specific. Similar to the observations of Muir and Martin are those of Neufeld and Hiine,25 which showed that yeast cells will absorb both alexin and opsonin out of serum. A further similarity between the two serum constituents is the fact that both are absent from the normal fluid of the anterior cham- ber of the eye, but they together appear in it after injury (puncture for the first removal of fluid). A like parallelism between the ab- sence and presence of both has been shown for edema fluids. Furthermore, phosphorus poisoning which reduces alexin likewise reduces opsonin. Although this parallelism is very striking, it does not on this account mean that necessarily the two are identical. It may signify merely that the alexin is a necessary participant in normal opsonic action, essential in that it activates a thermostable opsonic constitu- ent just as it activates hemolytic or bactericidal sensitizer. This opinion has been expressed by Levaditi, Neufeld,26 Dean,27 and others, and indeed it is a conception which seems most logical. For the procedures which remove both alexin and opsonin, as stated above, do not, as a matter of fact, remove all the opsonic action. (Although Neufeld maintains this.28) Studies of Hektoen and others have definitely proved that, though reduced to almost nil, nevertheless heated serum shows definite though slight opsonic action as compared with indifferent menstrua such as salt solution. A similar slight remnant of opsonic action after absorption of normal serum with sensitized cells, bacteria, and precipitates is evident in the protocols of Muir and Martin. The significance of this point be- comes immediately clear when we consider the properties of the bac- teriotropins or immune opsonins, which are heat stable and capable of initiating opsonic action in the entire absence of alexin or comple- ment. It is possible, therefore, that there may be present in normal serum a slight amount of specific thermostable opsonin, which, though capable of acting feebly by itself, is nevertheless powerfully activated by alexin — just as bactericidal or hemolytic antibody is similarly activated. One of the most thorough studies upon this question is that of 25 Neufeld and Hiine. Arb. a. d. kais. Gesundh. Ami., Vol. 25, 1907. 26Neufeld. "Kolle u. Wassermann's Handbuch," Erganzungsband 2, p. 313. 27 Dean. Brit. Med. Jour., 2, 1907, p. 1409. 28 In fact he states that heated normal serum may be used as a control in opsonic experiments instead of salt solution. v 1 318 INFECTION AND RESISTANCE Cowie and Chapin.29 Dean 30 had previously shown that, although heated immune serum was capable of exerting opsonic action by itself, this action could nevertheless be enhanced by the addition of a little diluted fresh normal serum. The particular significance of Dean's work will be discussed later. Cowie and Chapin, however, carried on similar experiments with normal serum in which they at- tempted to reactivate heated normal serum by the addition of small amounts of diluted fresh serum, by itself but slightly opsonic. One of their experiments may serve to illustrate this point, as follows : Experiment 10. June 18, 1907 Phagocytic count n 1. Unheated serum 15 . 44 2. Salt solution 0. 18 3. Heated serum, 57° C 1 .08 4. Diluted serum (1 :15) 1 .56 5. Heated serum 57° C. + diluted serum (1 :15) 12.40 6. Unheated serum -f- unheated serum 16 . 08 This experiment and others like it seem to demonstrate clearly that the opsonic action of normal serum, though dependent largely upon alexin, is nevertheless also dependent upon a heat-stable body, comparable to the sensitizer or amboceptor, in that it is reactivable to almost the full power of the original condition (before heating) by slight amounts of alexin — in themselves almost inactive.32 These findings were later confirmed by Eggers,34 and it is plain from this work that the apparent opsonic inactivation of normal serum by heat depends upon the destruction of the heat-sensitive constituent only — the heat-stable substance — surely involved in the process, remaining intact, and reactivable. Closely associated with this phase of the problem is that of the specificity of the normal opsonins. For if, as at first supposed, the normal opsonins are, like complement or alexin, non-specific, the above amboceptor-complement structure of this mechanism would be rendered unlikely. Earlier work upon this question was con- tradictory. Bulloch and Western,35 working with staphylococci and 29 Cowie and Chapin. Jour. Med. Res., Vol. 17, 1907, pp. 57, 95 and 213. 30 Dean. Loc. cit. 31 Phagocytic count = average number of bacteria in each leukocyte. 32 In earlier experiments Hektoen and Ruediger 33 did not succeed in reactivating heated sera and concluded that normal opsonins had the hypo- thetical structure of toxins in that they possessed a haptophore and an opsonophore group. From this point of view Hektoen has subsequently receded largely because of work done under his own direction. 33 Hektoen and Ruediger. Jour. Inf. Dis., 1905. 34 Eggers. Jour, of Inf. Dis., Vol. 5, 1908. 35 Bulloch and Western. Proc. Roy. Soc. B., 77, 1906. FACTORS DETERMINING PHAGOCYTOSIS 319 tubercle bacilli, found that each of these organisms absorbed out separately specific opsonins from normal serum, leaving those for other bacteria but slightly reduced. Slight reduction of the opsonic action for other micro-organisms might easily be explained by a partial removal of complement which is bound to take place in such experiments. Simon, Lamar and Bispham,36 and some others failed to find any such specificity. Russell,37 Axamit and Tsuda,38 and a number of others obtained similar negative results — in that a num- ber of different bacteria seemed to absorb opsonins out of normal serum indiscriminately and without specificity. On the other hand, more recent careful work by Rosenow,39 by Macdonald,40 and by Hektoen 41 has upheld the original contention of Bulloch and West- ern. The work of Rosenow, in which pneumococci were shown to absorb out their specific opsonins from normal human serum, taking out in part only those for streptococci, staphylococci, and tubercle bacilli, is especially convincing, and the experiment of Hektoen with normal hemopsonins (opsonins which cause the phagocytosis of red blood cells) bear him out. It seems fair to conclude, therefore, that normal opsonins de- pend upon the cooperation of a heat-stable and a heat-sensitive body. The heat-stable body, analogous to normal sensitizer or amboceptor, is specific and reactivable by the heat-sensitive body which appears to be identical with alexin. This statement merely asserts the facts of the dual mechanism of the process without assuming necessarily the identity of the heat-stable body with sensitizer or that of the heat- sensitive one with alexin, though this seems extremely probable. This question we will discuss again more particularly in connec- tion with the bacteriotropins or immune opsonins. Further proof for such a complex constitution of the normal opsonins has been adduced by means of absorption experiments at 0° C. — by Cowie and Chapin. In our discussion of the lytic anti- bodies we have seen that sensitizer or amboceptor may be absorbed from serum by its specific antigen at 0° C. — but that the attachment of alexin takes place only when the temperature is raised above this. Practically no alexin is bound at the low temperature. Cowie and Chapin, applying this method of investigation, showed : 1. That normal human serum may have its opsonic power for staphylococci removed by absorption with staphylococci at 0° C. 2. Serum so treated retains the power of reactivating the op- sonin of heated normal serum. 36 Simon, Lamar, and Bispham. Jour. Exp. Med., Vol. 8, 1906. 37 Russell. Johns Hopk. Bull, Vol. 18, 1907. 38 Axamit and Tsuda. Wien. klin. Woch., Vol. 20, No. 35, 1907. 39 Rosenow. Jour. Inf. Dis., Vol. 4, 1907. 40 Macdonald. Quoted from Hektoen, loc. cit.; Aberdeen Univ. Studies, Vol. 21, 1906, p. 323. 41 Hektoen. Journ. Jnf. Dis., Vol. 5, 1908. 320 INFECTION AND RESISTANCE 3. Staphylococci so treated are more easily subject to phago- cytosis in the presence of dilute normal serum, or normal serum which has been inactivated by contact with Staphylococci in the cold, than are the same bacteria untreated. Kurt Meyer 42 has carried out similar experiments with paraty- phoid bacilli and normal serum, and, though his work is less exten- sive, he reaches the same conclusion as Cowie and Chapin. We may accept, therefore, as fairly well established that the opsonic power of normal serum depends upon a complex mechan- ism consisting of (a) a thermostable substance comparable to amboceptor or sensitizer, probably specific, but present in very small amount, and (b) a thermolabile substance probably identical with alexin or complement which powerfully, but non-speci- fically, enhances the slight opsonic power of the thermostable substance. In considering this conception, together with the subsequent dis- cussion of bacteriotropins or immune opsonins, it will be well to remember that in normal inactivated sera the thermostable opsonic constituent differs in its action from the bodies we speak of as ambo- ceptors or sensitizers in that it may functionate for phagocytosis by itself — entirely without alexin — while neither bactericidal nor he- molytic effects can be brought about by sensitizer alone. Does this definitely exclude the identity of this thermostable opsonic substance and sensitizer ? It is indeed an argument against identification, but in opsonic action, we must remember, there is merely a sensitization to the action of the phagocyte. This phagocyte may in itself be capable of furnishing a small amount of substance comparable in action to alexin — in fact, we have seen that the origin of alexin from leukocytes is still suspected by a number of workers. At any rate the phagocyte is a living cell which may well be capable of supplying in itself to some degree the necessary activation, and therefore the difference cited above is not necessarily a proof that the normal thermostable opsonic constituent is different from normal sensitizer or amboceptor. The difference between the opsonic action of normal serum and that of immune serum, then, is the fact that heating to from 56° to 60° C. almost completely destroys the former, whereas it has but slight if any diminishing effect upon the latter. The immune op- sonins, or, as Neufeld and Eimpau have called them, bacteriotropins, therefore are thermostable. This was determined as early as 1902 by Sawtschenko,43 and was subsequently studied with great accuracy 42 Kurt Meyer. Berl. klin. Woch., 1908, p. 951. 43 Sawtschenko. Ann. de Vlnst. Past., Vol. 16, 1902, quoted from Levaditi. FACTORS DETERMINING PHAGOCYTOSIS by Neufeld and Rimpau,44 Neufeld and Topfer,45 Dean,46 Hektoen,47 and others. It was shown that when an animal is immunized with any given bacterium or other cellular antigen (blood corpuscles, etc.) opsonic substances specific for the particular antigen appear in considerable quantities, and these are but slightly, if at all, dimin- ished when the serum is heated ; Neuf eld and Hiine 48 found that heating for as long as three-quarters of an hour to 63° C. did not noticeably reduce the activity of the bacteriotropins of immune serum, and that, again, unlike the normal opsonins, prolonged pres- ervation, under sterile conditions, changes them but slowly. These facts alone indicate a close similarity between the bac- teriotropins and the other well-known thermostable constituents of immune sera, and the question here again immediately arises whether we are to regard them as identical with any of the other specific antibodies or as distinct substances independent of these. It was suggested early in these investigations by Muir and Mar- tin that bacteriotropins might be identified with agglutinins, inas- much as they possessed resistance to heat, were active without ap- parent dependence upon alexin, and could not, at least in the earlier studies, be reactivated by the addition of fresh normal serum when once inactivated. The supposition was that for this reason the bac- teriotropin might have a structure like the hypothetical "haptines of the second order77 which Ehrlich attributes to the agglutinins. This supposition has found no experimental support in that ag- glutination and bacteriotropic effects did not run parallel. We our- selves are not ready to admit that such lack of parallelism is proof against their identity. However, since it is very probable that both agglutination and precipitation are merely phenomena of colloidal flocculation effects which follow certain quantitatively adjusted com- binations of antigen and specific antibody, and that it is not at all necessary to assume separate agglutinating or precipitating serum constituents, this problem becomes merely another version of the question of the identity of bacteriotropins and sensitizer or ambo- ceptor. Apart from thermostability, further similarity lies in the fact that bacteriotropins are strictly specific and may be specifically ab- sorbed out of immune sera by their respective bacteria. Like amboceptor or sensitizer they are specifically increased to a powerful degree by the treatment of animals with any given micro- organism and may be incited not only by the injection of bacteria 44 Neufeld and Rimpau. Deutsche med. Woch., No. 40, 1904 ; Zeitschr. f. Hyg., 51, 1905. 45 Neufeld and Topfer. Centralbl. f. Bakt., 1, 38, 1905. 46 Dean. Proc. Eoy. Soc. B., 76, 1905. 47 Hektoen. Jour. Inf. Dis., 3, 1906, and loc. cit. 48 Neufeld arid Hiine. Arb. a. d. kais. Gesundh. Amt.} Vol. 25, 1907. INFECTION AND RESISTANCE but by that of blood cells as well. In spite of these points of likeness, however, Neufeld 49 and his associates maintain rigidly that the two substances are not the same and that the bacteriotropins are distinct and independent antibodies. Among the reasons advanced in support of this opinion are the facts that certain immune sera, both antibacterial and hemolytic, may contain bacteriotropins without containing lysins and vice versa. That this is undoubtedly true has been shown not only by Neufeld and his associates but by Hektoen 50 and others, and it is likewise a fact that in sera in which both functions are demonstrable they fre- quently do not run quantitatively parallel. These are unquestionably strong arguments, but their force is somewhat weakened, as Levaditi has pointed out, by the fact that there are many varieties of bacterial immune sera which undoubtedly sensitize the specific bacteria (as can be shown by alexin fixation), but which do not lead to bacterio- lysis. Wassermann 51 also attaches little value to the lack of parallel- ism between the lytic and opsonic functions, expressing the belief that the solubility of the particular antigen may determine whether sensitization leads to phagocytosis or to lysis. With bacteria like the cholera spirillum rapid lysis takes place, but when, as in pneumo- cocci or streptococci, there is great resistance to lysis, sensitization may lead to delayed lysis anticipated by leukocytic accumulation, phagocytosis, and intracellular digestion. It by no means follows from mere lack of parallelism, therefore, that the two serum functions are dependent upon separate antibodies, although the argument is sufficiently strong to impose conservatism in identifying them. Another important argument advanced against the identification of bacteriotropins with the bactericidal sensitizers or amboceptors is the fact that the former lead to phagocytosis without the participa- tion of alexin, whereas the latter become active for lysis only when alexin is present. This point also has constituted Neuf eld's strongest support for maintaining that the bacteriotropins or immune opsonins are entirely distinct from the normal opsonins. It is true, indeed, that immune serum, unlike normal serum, may opsonize powerfully even after heating to temperatures which destroy alexin. If we regard the heat-stable lytic antibody as an amboceptor in the strict sense of Ehrlich, as a specific "Zwischenkb'rper" with a complementophile group, this argument would have considerable weight. Even in this case, however, strong sensitization of the bac- teria may make them amenable to the living cells — the phagocytes — *9Neufeld and Topfer. CentralU. f. Bakt., 1, 38, 1905. 50 Hektoen. Jour, of Inf. Dis., 6, 1909. 51 Wassermann. Deutsche med. Woch., Vol. 33, No. 47, 1907. FACTORS DETERMINING PHAGOCYTOSIS 323 which in itself may furnish a slight amount of alexin or alexin-like substances. We may regard the action of the immune serum upon the antigen as rather a sensitization in the sense of Bordet, and it does not seem logical to assume that the heat-stable bodies, similar in other respects, are different merely because they can sensitize bacteria both to the action of an alexin and to that of a living cell, which in itself surely contains a number of different enzymes, comparable functionally to alexin, though possibly not identical with it. Indeed, the experiments of Dean have given much positive evi- dence in favor of regarding the immune opsonins or bacteriotropins as true amboceptors or sensitizers. Dean 52 found that, although heated immune serum may unquestionably opsonize by itself, its action may be still further enhanced by the addition of a little diluted normal serum (compare these results with those of Cowie and Chapin on normal opsonins). Hektoen's 53 experiments with hemopsonic immune sera are analogous. We cite one of these as illustrating the point in question : TABLE I Phagocytosis of Goat Corpuscles under the Influence of Goat-blood-immune Rabbit Serum, and Normal Guinea Pig Complement (Table from Hektoen, loc. cit.) Immune serum Normal guinea pig serum Phagocytosis .001 .001 -+ .01 .01 4. 20. 0 Here, therefore, the diluted immune serum, but slightly cyto- tropic in itself, was powerfully activated by a diluted unheated nor- mal serum, which in itself was entirely inactive. Indeed, an experiment by Neufeld himself, with Bickel,54 points in the same direction. They found that, when a heated specific hemo- lytic serum was added to the homologous cells in such small quanti- ties that it no longer exerted cytotropic (opsonic) action, the addi- tion of a small amount of alexin, too small to lead to hemolysis of the cells (and not by itself cytotropic or hemopsonic), caused active phagocytosis. Analogous experiments upon bacterial antisera were carried out by Levaditi and Inmann. It thus appears that, even in the case of the immune opsonins or bacteriotropins, we may think of a participation of two substances — a sensitizer-like one and one com- parable to alexin or complement. We may, at least, infer that the full opsonic action both of normal and immune sera is dependent 52 Dean. Proc. Royal Soc. B., 79, 1907. 53 Hektoen. Jour. Inf. Dis., Vol. 6, 1909, p. 67. 54 Neufeld and Bickel. Arb. a. d. kais. Gesundh. Amt., Vol. 27, 1907. 324 INFECTION AND RESISTANCE upon the cooperation of two such bodies. It is likely, therefore, that the mechanism of normal and of immune opsonic action may, after all, differ only in quantitative relations between the two. For assuming this to be an antibody-alexin mechanism like hemol- ysis, we may recall the work of Morgenroth and Sachs on the rela- tions between amboceptor and complement in hemolysis. There we saw that a large amount of amboceptor would cause hemolysis in the presence of a small amount of complement and vice versa. There- fore, here, too, in normal serum the small quantity of amboceptor or specific thermostable opsonin (bacteriotropin) may act very power- fully in the presence of the alexin. When the latter is destroyed, however, the minute quantity of specific thermostable opsonin is hardly enough to do more than initiate slight phagocytosis of com- paratively non-resistant bacteria, whereas the large amount of spe- cific sensitizer left in immune sera after inactivation may still lead to strong bacteriotropic action. In outlining this explanation we have consistently drawn upon the analogy between thermostable op- sonin and amboceptor or sensitizer. Whether or not these two sub- stances are identical is by no means positively determined and must be considered an open question for the present. However, from the above, it seems to us that much testifies in favor of such an identi- fication.55 The preceding discussions have ignored the possibility that apart from opsonic or bacteriotropic action on the bacteria there may be a difference in phagocytic energy which depends upon inherent prop- erties of the leukocyte itself. Indeed, the technique by which the researches of Wright and his followers were carried out does not in any way take into account the source of the leukocytes as a possible variable factor. There is, however, a considerable amount of evidence which points to differ- ences in phagocytic powers residing in the leukocytes themselves independent of the serum. Park and Biggs 56 have demonstrated such differences for the leukocytes of normal persons in the phagocy- tosis of staphylococci, and more extensive researches have been made with similar results, in the case both of staphylococci and tubercle bacilli by Glynn and Cox.57 The last-named authors, moreover, recognized the necessity, in making such investigations, of experimenting with leukocyte emul- sions containing approximately the same number of cells, for, as Fleming 58 had shown, if unequal leukocytic emulsions are used, 55 Pf eiffer (quoted from P. Th. Muller) regards opsonic action as due to a combined action of amboceptor and complement and speaks of it as an "Andauung" of the bacteria for the leukocyte — which we may translate best as a partial predigestion. 56 Park and Biggs. Jour. Med. Res., Vol. 17, 1907. 57 Glynn and Cox. Jour. Path, and Bact., 14, 1910. 58 Fleming. Practitioner, London, Vol. 80, 1908. FACTORS DETERMINING PHAGOCYTOSIS 325 less phagocytosis per cell occurs in the emulsion containing the greater number of leukocytes. This phase of the subject has been taken up most thoroughly by Hektoen 59 and his associates, and Rose- now 60 has made careful comparative studies on pneumococcus phagocytosis, in which he standardized the leukocytic suspensions by actual cell counts. His work as well as that of Tunnicliff,61 of the same school, has shown definitely that the inherent phagocytic power of leukocytes may vary not only in health and disease, but differences may exist between the cells of apparently normal people. Tunni- cliff showed, for instance, that at birth the leukocytes are less active than in adult life. For accurate experimental work, therefore, as well as in theoret- ical reasoning upon problems of phagocytosis, it is necessary to bear in mind the possible inherent variations in the leukocytes themselves. Of the three factors concerned in the process of phagocytosis, then, we have considered two, the serum and the leukocytes. The former we have seen exerts a powerful determinative influence on the process, the latter a less marked influence, though still definite and measurable. We have still to discuss the bacteria themselves as variable factors in determining the degree to which phagocytosis may take place. This problem was first investigated by Denys and Marchand in connection with their work upon streptococcus immunity, and was further studied in detail by Marchand. Marchand 62 showed that leukocytes would readily take up non-virulent streptococci in the presence of normal serum, but that under similar conditions virulent streptococci were not phagocyted at all or to a very slight degree only. He determined further that this resistance to phagocytosis remained unchanged after the virulent organisms had been killed by heat, and washed clean of culture fluid. It seemed, therefore, that the resistance depended upon a condition of the bacterial body and not upon substances secreted and given off to the environment. These experiments, as well as similar work by Mennes,63 Gruber and Futaki,64 and others makes it clear that differences in virulence between different species of bacteria, as well as between different strains of the same micro-organism, depend, at least in part, upon the resistance which the bacterial bodies oppose to ingestion by the leukocytes. We must distinguish clearly here between these appar- ently purely "antiopsonic" bacterial properties and those supposedly "antichemotactic" substances which are conceived as a cause for 59 Hektoen. Jour, of A. M. A., Vol. 57, No. 20, 1911. 60Rosenow. Jour, of Inf. Dis.y 7, 1910. 61 Tunnicliff. Jour. Inf. Dis., 8, 1911. 62 Marchand. Arch, de Med. Exp., No. 2, 1898. 63 Mennes. Zeitschr. f. Hyg., Vol. 25, 1897. 64 Gruber and Futaki. Munch, med. Woch., 1906. INFECTION AND RESISTANCE virulence by Deutsch and Feistmantel 65 and by Bail 66 in his so- called "aggressins." The latter are supposed to be secreted bacterial substances by means of which the leukocytes are heid at bay. The properties we are, at present, considering are probably in no way antichemotactic, but oppose purely the actual ingestion by the leukocyte, nor do they seem to depend upon the secretion of sub- stances which injure the leukocytes. For, in the first place, a profuse accumulation of leukocytes may follow the injection of virulent micro-organisms, and Denys (quoting from Gruber) has seen active phagocytosis of virulent pneumococci, but none of virulent streptococci when antipneumococcus serum was injected with the mixture. Rosenow 6T has carried out a thorough investigation dealing with these relations in pneumococcus infection. Seventy-five strains of this organism were all found non-phagocytable when first isolated and the resistant condition was associated with virulence for rabbits and guinea pigs. It was found, moreover, that the resistance to phagocytosis was dependent upon the inability to absorb opsonin. For, while phagocytable non-virulent pneumococci absorbed specific opsonin from serum, the virulent ones failed to do this in proportion to the degree of their virulence. Furthermore, extraction of the bodies of the virulent organisms in NaCl solution yielded a substance which inhibited the action of pneumococcus opsonin — a true anti- opsonin — which he speaks of as "virulin." This discovery, if con- firmed, would supply us with a very simple explanation for some phases of the problem of virulence. It is, indeed, likely that the antiopsonic property is closely bound up with chemical and struc- tural changes which take place in the bacterial cell as it adapts itself to the parasitic conditions. This is plain from the fact that pneumo- cocci and some other bacteria will rapidly lose their virulence when cultivated on artificial media devoid of animal serum, will retain it longer if grown on some serum media, and will rapidly regain it if passed through animals. The formation of a capsule is unquestion- ably a morphological evidence of such a change. Habitually capsu- lated bacteria, like the Friedlander bacillus, and Streptococcus muco- sus, are of fairly constant virulence, while in other micro-organisms like the pneumococci, anthrax bacillus, plague bacillus, and certain other streptococci, the formation of a capsule goes hand in hand with an increase of virulence. By the aid of this morphological earmark of virulence, moreover, Gruber and Futaki have obtained further 65 Deutsch and Feistmantel. Quoted from Sauerbeck. Lubarsch und Ostertag, Vol. 2, 1906. 66 Bail. Arch. f. Hyg., Vol. 52, 1905. 67 Rosenow. Jour. Inf. Dis., Vol. 4, 1907. FACTORS DETERMINING PHAGOCYTOSIS 327 proof that the resistance to phagocytosis in these cases is due to the nature of the bacterial cell body rather than to any secreted anti- opsonic substances. For, after the injection of anthrax bacilli into guinea pigs, they saw that leukocytes would take up uncapsulated bacilli, apparently picking them out of the midst of surrounding capsulated organisms which they were unable to ingest. CHAPTER XIV THE OPSONIC INDEX AND VACCINE THEEAPY WRIGHT'S 1 investigations upon phagocytosis were, indirectly, the outcome of his earlier work upon antityphoid vaccination. His purpose .in these studies had been a purely practical one, and he had attempted to obtain a guide for the dosage and the interval be- tween injections by measuring the bactericidal and agglutinating powers of the blood serum. In the case of typhoid immunization this was indeed a practicable method of control, since the bacteri- cidal power of the blood serum rose directly as the immunization of the patient was attained. In the cases of many other bacteria, how- ever, this method of study was not practicable, and Wright, as others before him, did not find a regularly increased specific bactericidal power in the blood sera of immunized animals or of patients con- valescing from infections with such bacteria as the staphylococcus, streptococcus, Micrococcus melitensis., the Bacillus pesiis, and a num- ber of others. In fact, together with Windsor,2 he showed that nor- mal human blood has practically no bactericidal power for pyogenic staphylococci and that antistaphylococcus inoculations or recovery from an infection do not result in the production of such proper- ties in the serum. These determinations are practically identical with Nuttall's 3 earlier studies on the same bacteria and, indeed, cor- respond with the data obtained by Metchnikoif and his followers in their work on anthrax infection. For, in discussing these investi- gations, we saw that very often the serum of a comparatively resist- ant animal is less potently bactericidal than that of a more suscep- tible one. We need only recall the difference between rabbits and dogs in this respect. The serum of the former is more strongly bac- tericidal than that of the latter, and yet rabbits are the far more susceptible animals. These relations have been studied with great care, also, by Petterson.4 It was logical in such cases to look for the cause of resistance in the activity of the phagocytes, and this, we have seen, Metchnikoff did successfully in a large series of cases, both as regards natural and acquired immunity. 1 Wright. Lancet, 1902; Practitioner, Vol. 72, 1904. 2 Wright and Windsor. Jour, of Hyg., Vol. 2, 1902; and Wright, Lancet, 1900 and 1901. 3 Nuttall. Zeitschr. f. Hyg., Vol. 4, 1888. 4 Petterson. Centralbl. f. Bakt., Vol. 39. 328 OPSONIC INDEX AND VACCINE THERAPY 329 Yet the controversy between the strictly humoral and the cellular schools was by no means regarded as closed, especially since, in such cases as typhoid infection, the parallelism between increased resist- ance and extracellular bactericidal power was so plainly evident, while in this disease particularly (for technical reasons which will become clear as we proceed) no such parallelism with phagocytosis could at first be shown. It was because of such apparent confusion that Leishmann5 undertook to study again the relation of phago- cytosis to active immunity, chiefly upon staphylococcus cases that were being "vaccinated" therapeutically by Wright himself. In order to obtain a numerical measure of the degree of phago- cytosis, he developed a simple technique which, though crude, served to give him the information he sought. It consisted in taking small quantities of the blood of patients and mixing these in capillary pipettes with equal volumes of bacteria suspended in salt solution. The mixtures were then placed on slides, covered with a cover- slip, and incubated at 37° C. for varying periods. At the end of incubation the preparations were smeared upon slides and stained by Leishmann's modification of the Romanowski method, the num- ber of bacteria in a large series of leukocytes counted and an average taken. This method had many serious flaws, chief among them being the liability to coagulation of the preparations and the fact that, in each test, the fluid constituents as well as phagocytes, both of them variable factors, came from .the same individual. While, therefore, it was possible to estimate an increase or decrease of general phago- cytic power, it was impossible to analyze this in reference to its de- pendence either upon the condition of the cells, on the one hand, or that of the plasma or serum, on the other. Moreover the relation of the number of leukocytes to that of bacteria in individual tests neces- sarily differed, and this, we have seen, adds a variable factor which renders it impossible to compare any two experiments with accuracy. In spite of these difficulties, however, Leishmann succeeded in establishing, in a number of cases of staphylococcus infection, that an increased resistance was accompanied by an increased energy of phagocytosis. Leishmann, however, went no further than this, and interpreted his results on the basis of the "stimulin" theory of Metchnikoff. The subsequent studies of Wright, which began at the point at which Leishmann stopped, have been described in the preceding chap- ter and had, as their main result, we have seen, the discovery of the opsonins and the final confirmation of Denys' conception of the true mechanism of cooperation between serum and leukocytes in phago- cytosis. In order to carry out these studies the technique of Leish- 5 Leishmann. Br. Med. Jour., 1, 1902; Transact. Lond. Path. Soc., Vol. 56, 1905. 330 INFECTION AND RESISTANCE mann was quite inadequate, and Wright's first task was to modify it in such a way that reasonably accurate comparative estimates of phagocytosis could be made. It is necessary to outline Wright's method briefly in this place in order that we may consider possible sources of error and obtain a clear understanding of the conclu- sions he based on his observations. Wright recognized that the deter- mination of the degree of phago- cytosis, induced by the opsonin of WRIGHT CAPSULE FOR TAKING any given serum in a ^ single test, is BLOOD TO OBTAIN SERUM FOR by itsell o± no value, since the actual OPSONIC TESTS. number of bacteria taken up by each leukocyte, apart from the opsonic contents of the serum, depends also upon such purely technical fac- tors as the concentration of the bacterial emulsion, the relative num- ber of leukocytes, and the length of time of incubation. Two individual tests, therefore, carried out with the serum of the same patient at the same or at different times, with different bacterial emulsions or leukocytes in each, would give variable results, even though the opsonin contents them- selves were entirely alike. In order, therefore, to obtain 9 relative estimate of the opsonic contents of any serum it is neces- sary to compare the phagocytic ac- tivity induced by this serum with the similar power of another sup- posedly normal serum, both tests being carried out, under exactly similar conditions, with the same bacterial emulsion and the same leukocytes. The average number of bacteria found in each leuko- cyte in each one of the prepara- tions is then the "phagocytic in- dex." The relation of the phago- cytic index of the unknown serum METHOD to that of the supposedly normal serum constitutes what Wright has called the "opsonic index." Instead of using the whole blood of the patient Wright takes a small amount of blood in glass capsules, allows it to clot, and uses the expressed serum in his test. For comparison with this he em- ploys a "pool" of a number of specimens of serum from supposedly OF PRODUCING AN EVEN EMULSION OF BACTERIA FOR OP- SONIN DETERMINATION. OPSONIC INDEX AND VACCINE THERAPY 331 normal individuals. By the use of such a serum mixture any slight possible variations from the normal in any one of the sera are likely to be equalized, and a closer approach to a normal standard is at- tained. The leukocytes used in both tests are the same and taken, as a rule, from the blood of the worker or from some other supposedly healthy person. They are obtained by taking 15 or 20 drops of blood from the finger or ear into 5 to 10 c. c. of sodium citrate solu- tion, in which the blood does not clot. Brief centrifugalization throws down the blood cells, with a thin, buffy coat of leukocytes on top, and these are gently taken off with a pipette. This constitutes the leu- kocytic cream of Wright's experiments, and furnishes a uniform leu- METHOD OF TAKING UP EQUAL VOLUMES OF LEUKOCYTES, BLOOD SERUM AND BACTERIAL EMULSION IN WRIGHT'S TECHNIQUE FOR OPSONIC-INDEX DETER- MINATION. kocyte factor for the two tests which are to be compared. The bac- teria are obtained by emulsifying carefully in salt solution. It is very important to obtain an emulsion free from clumps and neither too thick nor too thin, a result which can be secured only by experi- ence. Equal quantities of serum (unknown and normal "pool" respec- tively) are mixed with equal quantities of the bacterial emulsion and the leukocytes in capillary pipettes, and the mixtures are incubated for fifteen to thirty minutes under exactly similar conditions. At the end of this time smears are made upon slides, the preparations stained, and the numbers of bacteria in a hundred or more leuko- cytes counted in each of the two experiments. The average is taken, and from the phagocytic indices thus obtained the opsonic index is calculated. For instance, if Phagocytic index (normal pool) = 8 Phagocytic index (patient's serum) = 6 then the opsonic index (patient's serum) = 0.75. Or, if the phago- cytic index of the normal pool had been 10. and that of the patient's serum 15., then the opsonic index of the patient's serum, higher than normal, would be 1.5. For the insurance of accuracy in carrying out this method Wright calls especial attention to the caliber of the capillary pipettes that are used, the concentration of the sodium citrate solution, which should be 1.5 per cent., and the freshness of the leukocytes. But it is still necessary to remember that with the greatest care in tech- INFECTION AND RESISTANCE nique uncontrollable sources of error influence this method. Most important among them are the differences necessarily existing be- tween different normal sera used for comparison and differences in the agglutinative powers of the sera used in the two specimens. For it is plain that different degrees of agglutination may bring about great variations in the number of bacteria with which the individual leukocyte comes into contact. Wright's method has also been particularly unsatisfactory in taking the opsonic index against such bacteria as the typhoid bacillus and the cholera spirillum, or- ganisms which are very rap- idly digested after being taken up by the leukocytes. In consequence, even after as short an incubation time as five or ten minutes, the in- gested bacteria are partly dis- integrated, are stained in- distinctly, and cannot be counted with accuracy. In order to avoid this source of error Klien 6 has devised a modification which depends upon gradual dilution of the serum in a series of pha- gocytic tests with the LEUKOCYTES CONTAINING BACTERIA. DRAW- same leukocytic and bacterial ING OF FIELD AS SEEN IN WRIGHT'S emulsions. In this way he METHOD OF OPSONIC-INDEX ESTIMATION. , J . ,11 f ••• determines tne degree 01 di- lution of the serum to be tested at which phagocytosis no longer exceeds that taking place in salt solution alone. The degree of dilution at which this result was obtained has been called by Simon the "coefficient of extinction." A comparison of sera with regard to this value, it is clear, furnished an estimate of their quantitative opsonic properties quite as instructive as the direct estimations by the Wright method, and in our opinion, at least, more reliable. Though also subject to some of the objections advanced against the Wright method, it has the definite advantages mentioned above, and is not so closely dependent upon irregularities in counting, agglutinin influences, and differences in relative pro- portions of bacteria and leukocytes employed. Jobling7 has used this method with success for the standardization of antimeningitis serum. A further modification suggested by Simon, Lamar, and Bis- 6 Klien. Johns Hop. Hosp. Bull., Vol. 18, 1907. 7 Jobling. "Studies from the Rockefeller Inst.," Vol. 10, 1910, p: 614. OPSONIC INDEX AND VACCINE THERAPY 333 pham 8 depends upon a combination of the dilution method and a modification in the method of counting. They make comparative tests of the same serum, diluted from 1 to ten to 1 to one hundred in salt solution, and estimate the opsonic power, not by determining the average number of bacteria to the leukocyte, but by taking a per- centage of the total number of leukocytes which take part in the phagocytosis, that is, contain any leukocytes at all. The bacterial emulsion for this method should be so thin that, in normal serum, only about 50 per cent, of the leukocytes will contain bacteria. That Wright's method, or any of the others, gives absolutely accurate results will hardly be claimed by any one who has worked upon opsonic-index estimations. There are certain uncontrollable variable factors, some of which have been pointed out above; and, apart from these, the delicacy of the technique is such that reliable results can ordinarily be obtained only by trained workers after con- siderable practice and experience. Even in such hands the per- centage of personal error is more likely to be above than below 10 per cent. For ordinary clinical purposes, therefore, in the control of cases the estimation of the opsonic index is not often practicable. On the other hand, there can be little doubt about the fact that careful comparative estimation, by Wright's method and by some of the modifications, carried out by workers with experimental train- ing and consequent attention to extensive controls, have yielded results of sufficient accuracy to permit the recognition of definite facts concerning opsonins. It is beyond question, therefore, that the conclusion regarding the relation of opsonic fluctuations to clinical conditions and the general significance of opsonins emanating from laboratories like those of Wright, Neufeld, Hektoen, and some others may be accepted as fact — especially since in most essentials such workers have agreed. In consequence we are now in possession of knowledge regarding the opsonic constituents of the blood in health and disease, and in the course of active immunization wit\ bacterial vaccines, which is of the greatest practical importance. We may summarize the results of such investigations by saying that in many of the infections of man the resistance of the patient is roughly pro- portionate to the opsonic index — and that properly spaced inocula- tion with suitable quantities of dead bacteria (vaccines) will raise the opsonic index and lead to recovery in many of the localized subacute and chronic conditions. As to the usefulness of the treatment in various infections and the limitations within which we may hope for results opinions differ, and these will be discussed more fully below. Before we proceed to this, however, it will be useful to consider the studies upon which 8 Simon and Lamar. Johns Hop. Hosp. Bull, Vol. 17, 1906 ; Simon, Lamar, and Bispham, Jour. Exp. Med., Vol. 8, 1906; Simon, Jour. A. M. A., Vol. 48, 1907, p. 139. INFECTION AND RESISTANCE the parallelism between opsonic index and: clinical condition was founded. Wright's own earlier studies were made chiefly upon staphylococ- cus infections and tuberculosis. Since then the method has been applied to almost all known infections with varyingly successful results. One of the first steps in determining such a parallelism between the resistance of a patient and the opsonic index consisted, of course, in comparing the index of the sera of normal individuals with that of patients suffering from infection. Wright and Douglas did this in a large series of studies. In the case of staphylococcus infections the following experiment will illustrate their results: TABLE I (Wright and Douglas, Proc. Royal Soc., Vol. 74, 1904.) Showing the ratio in which the phagocytic or opsonic power of the patient's blood stood in each case to the phagocytic or opsonic power of the normal individual who furnished the control blood. (The phagocytic power of the control blood is taken in each case as unity.) Initials of Patient Form of Staphylococcus Invasion Opsonic Index E. E. Furunculosis 0 48 F. F. Sycosis 0.49 J. E. Acne 0 64 J. H. Furunculosis 0 87 W. B. Acne 0 55 E. H. Acne 0 82 W. H. Furunculosis 0.79 R. G. Furunculosis 0 7 G. L. Acne and sycosis 0 74 S C Furuncu losis 0 87 W. L. Furunculosis 0 88 W P Furunculosis 0 39 S. F. Very aggravated sycosis 0 1 E. F. D. Acne 0.73 D C Sycosis 0 8 J. M. Acne 0 48 W. M. Sycosis 0.37 E P Acne 0 6 M. S. Pustular affection of lips 0.6 F V. Repeated staph infection 0 47 In this series, as in others investigated by Wright and his col- laborators, staphylococcus infection was uniformly associated with a low index. He concludes that there is probably a causative rela- tion between the two facts, in that under conditions of depressed phagocytic powers staphylococci may gain a foothold, while under OPSONIC INDEX AND VACCINE THERAPY 335 ordinary normal conditions they would fall prey to phagocytic de- struction soon after entering the body. The study of the opsonic index during the treatment of such cases with dead staphylococcus cultures (usually with the organisms cultivated from the patient's own lesions — " autogenous vaccines") revealed a striking coincidence between the rise of the opsonic index and improvement in the clinical conditions. A number of further interesting and practically important points were brought out by the systematic study of these relations which may be illustrated by 769 JO//IZ/3/4/S/6/7M/9&&&8&&& 2.0 /.9 1.6 /.7 /6 /5 /* /3 12 /./ 10 .9 8 7 .6 6 ^ X ^ f \ J ¥ 1 \ w) (} k 1 \ 8 k «Q 1 \ s? \ \ Mi 1 4 Q 5J $ 1 | 1 /\ < ' / ^ n k. / \ § I 0 (V / / A s / \ / f s * V * v \ IS <^ £ has made similar studies and claims to have found toxic protein cleavage products similar to his kenotoxin in exposed air. 10 See Zinsser and Young, Jour. Exp. Med., Vol. 17, 1913. 430 INFECTION AND RESISTANCE easily disposed of, would, on the basis of the preceding assumptions, result in a very gradual antigen destruction with consequent antibody formation, so that, at the end of eight to ten days, there would be present side by side remnants of unchanged antigen and newly formed specific antibody. The destroyed antigen fraction, in other words, gradually sensitizes the body to the fraction which persists and has not yet been assimilated or excreted at the end of this time. Such a point of view would explain, not only the reaction after a first in- jection, but would account for the incubation time in such cases, and for the differences between these reactions and both the "immediate" and the "accelerated" reactions of cases twice injected. The bearing which this point of view would have on the problems of incubation time in general is obvious. The recognition of the anaphylactic nature of serum sickness has led to many attempts to develop methods of antitoxin adminis- tration by which these reactions could be avoided. Since it was de- termined that the degree of reaction was directly dependent upon the amount of the foreign serum injected, it was an obviously logical procedure to attempt in antitoxin production to concentrate as high a potency of antitoxin into as small as possible an amount of serum. Attempts have also been made to alter the serum itself in such a way that it would lose its properties of acting as an anaphylactic antigen without suffering materially in antitoxin value, ^-Bujwid11 found that serum sickness was less frequent after the use of sera which had been allowed to stand for prolonged periods, and we have seen that Besredka and others have claimed a reduction of toxic property in sera heated repeatedly to 60° C. It was hoped, moreover, that the so-called concentration methods — such as those of Gibson, Banz- haf, and others — would yield an antitoxin that would be devoid of anaphylactic properties. None of these methods of altering the serum can, however, be said to have been satisfactory in that the antitoxic property seems to be closely associated with the globulins,12 which we have seen are at the same time closely associated with the production of anaphylaxis. The conclusions of Rosenau and Anderson 13 regarding this are based on direct experimentation with concentrated antitoxin made at the New York Department of Health by the Gibson method. They found the refined antitoxin, volume for volume, quite as toxic as the unrefined, but since the same amount of antitoxin is by this and other methods concentrated in a considerably smaller amount of 11 Bujwid. Quoted from Friedberger and Mita, Deutsche med. Woch.f No. 5, 1912. 12 Among others previously mentioned see also Turro and Gonzales, C. R. de la Soc. de BioL, Vol. 69, 1910. 13 Rosenau and Anderson. U. S. Pub. Health and M. H. S. Hyg. Lab. Bull. 36, April, 1907. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 431 protein solution there is a distinct gain for safety in the use of such preparations. Endeavors to produce potent antitoxic sera by chemical or physical methods without any sensitizing properties have thus been unsuccessful. On the other hand, the knowledge gained by animal experimen- tation regarding the influence upon the anaphylactic manifestations exerted by various methods of administering the antigen has led to results which have proven of much value, both in the immuniza- tion of experimental animals and in human serum therapy. Prob- ably the most carefully studied of these methods is the one which Besredka 14 has recommended on the basis of his work on antianaphy- laxis in animals. He found that sensitized guinea pigs could be injected with quantities of serum amounting to about one half or less of the fatal dose without showing symptoms, and subsequently, at intervals of 2 to 5 minutes, further injections of the serum could be given, the total amount five to twenty times exceeding the lethal dose without causing symptoms of any kind. From these experi- ments he has developed a method of serum injections the principle of which is very simply a division of dose. In lieu of injecting into an animal or man the entire quantity of serum at once, small, gradu- ally increasing amounts are administered in two, three, or more doses, the intervals varying from five minutes to several hours, ac- cording to the necessities of speed indicated by clinical considera- tions. The process as applied to man consists, then, in preceding the injection of the larger quantity of the serum by one or two sub- cutaneous injections of smaller amounts. With this principle well defined it would be quite unwise to lay down definite rules of quan- tity or interval at present, since in no instance will it be possible to estimate the exact condition of susceptibility of the particular case. It goes without saying that the precautions should be par- ticularly respected in children in whom the relation of 5 or 10 c. c. of serum volumes to the body weight approaches the dangerous pro- portions dealt with in animal experiments. Besredka has also shown that if the rectum of a sensitive animal is cleaned out by enema, and a relatively large amount of the an- tigen then introduced, an injection may be given in within twelve to twenty-four hours later without danger, however delicate the hyper- susceptibility of the animal has been. This method apparently must depend upon a slow, gradual absorption of antigen, and would seem to furnish a most convenient and advisable method to apply in man. 14 Besredka. Ann. de I'Inst. Past., Vol. 24, 1910; C. E. de la Soc. de BioL, 65, 1908, p. 478; C. E. de la Soc. de BioL, Vol. 66, 1909, p. 125; ibid., 67, 1909, p. 266; C. E. de I'Acad, des Sc., Vol. 150, 1910, p. 1456; ref. Bun. de I'Inst. Past., Vol. 8, 1910, p. 735. 432 INFECTION AND RESISTANCE Friedberger and Mita 15 have suggested another method which also depends upon very slow administration rather than division of dose. In experiments upon guinea pigs they had found that sensi- tized animals which, as tested by controls, would succumb to intrave- nous injections of 0.01 c. c. of sheep serum per 100 grams weight when the entire quantity was injected within one minute, would sur- vive a similar administration of as much as 0.1 c. c. if, by means of a specially constructed apparatus, the injection was made gradually, extending over a period of 100 minutes. While this method offers many practical difficulties to ordinary bedside application, it does show that the intervals of injections by the Besredka method do not need to exceed fractions of an hour — or, at most, a few hours — in order to add materially to the safety of injection. There is another phase of specific therapy in which the question of possible anaphylaxis must be taken into consideration, and that is the treatment of patients with bacterial vaccines. As a matter of fact the danger of anaphylaxis in such cases is probably very remote — both because of the shortness of the intervals at which these injec- tions are usually made and because of the extremely small amounts of protein represented by the usual dose of 100 or 200 millions of bacteria. However, the possibility cannot be disregarded, especially in children, and two cases were verbally described to the writer by Dr. Philip Van Ingen, in which gonococcus vaccines caused immedi- ate symptoms of such a character that anaphylaxis could not be ex- cluded. Ohlmacher16 also has described localized reactions at the place of inoculation as well as swelling and tenderness at points of former inoculations following bacterial vaccine injections. He has oc- casionally seen slight systemic symptoms (dizziness, nausea, etc.) which he explains on the basis of anaphylaxis. Moreover, it must be remembered that active sensitization with bacterial antigens has been most regularly successful in the hands of Kraus and Doerr,17 Holobut,18 and Kraus and Admiradzibi,19 as well as in confirmatory experiments carried out in the Stanford laboratory, when repeated injections at short intervals were made, rather than when, as in serum anaphylaxis, a single injection only was given. This would lend an even closer analogy to the procedures carried out during vaccine treatment. For instance, in the successful experiments of the last-named writers ten daily injections of 1/100 of a slant culture of dead colon bacilli were made for the purpose of 15 Friedberger and Mita. Deutsche med. Woch., No. 5, 1912; figures taken from Versuch., 3. 16 Ohlmacher. Jour. Med. Ees., Vol. 19, 1908, p. 113. 17 Kraus u. Doerr. Wien. klin. Woch., No. 28, 1908. 18 Holobut. Zeitschr. f. Immunitatsforsch., Vol. 3, 1909. 19 Kraus u. Admiradzibi. Zeitschr. f. Immunitatsforsch., Vol. 4, 1910. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 433 sensitization — the toxic dose of 1/2 slant being given fifteen days after the last of these. In order to obtain some opinion regarding the possible dangers of vaccine therapy in this regard the writer a few years ago ob- served carefully a pair of young goats, animals extremely favorable for anaphylactic experiment (and of approximately the weight of a child of three) in the course of frequent and irregularly spaced intravenous injections of typhoid bacilli. In both cases marked anaphylactic symptoms were observed after the animals had attained a considerable agglutinative and bactericidal power (1 to 5,000 to 1 to 20,000), but in each case only after intravenous injections of large quantities of bacteria, 1/2 to 2 slant cultures. While, of course, such experiments are not conclusive in any way, from these, as well as from a number of laboratory accidents in the course of animal immunization, it is the writer's impression that the intrave- nous injection of bacteria or bacterial products in human beings would be a procedure involving some risk, unless more thorough ex- perimental data than we at present possess were available to guide us as to dosage and intervals. The ordinary subcutaneous treatment of patients, however, with bacteria in the amounts customarily em- ployed in vaccines would seem to be practically without risk as far as acute anaphylaxis is concerned. In the treatment of animals with vaccines of various kinds Le- clainche2021 has repeatedly called attention to the fact that inocu- lation with a vaccine may lead to a condition of hypersusceptibility, serving to light up a latent lesion which might have been held in check if the normal resistance had not been interfered with. This objection, we have seen, has been made on numerous occasions against tuberculin therapy, and is one of the factors which have led to the great caution in dosage and control of all therapy based on active immunization. These considerations, even more than the rather remote dangers of serious active anaphylaxis, require that all forms of specific therapy should be carried out only under the safeguards of thorough familiarity with the experimental phases of such work. Our own recent studies on anaphylatoxins, moreover, have in- clined us to believe that hypersusceptibility to bacterial protein may well be a strong predisposing factor in infection. Serum sickness, occurring as a direct consequence of the injection of a foreign protein into a human being, forces itself upon us as manifestly related to anaphylaxis. There are a number of other clinical conditions which are less obviously anaphylactic in nature, but in which we have many good reasons for attributing an important part of the etiology to a state of hypersusceptibility. Thus the pe- 20Leclainche and Vallee. Ann. de I'Inst. Past., 1902. 21 Leclainche. Revue Gen. Med. Vet., Sept., 1911; Bull, de I'Inst. Past., 9, 1911, p. 1089. 434 INFECTION AND RESISTANCE culiar so-called "idiosyncrasies" observed in many people who suffer from urticarial skin rashes, gastro-intestinal difficulties, and even severe systemic illnesses after certain varieties of food seem to de- pend upon an acquired or possibly inherited hypersusceptibility to the particular proteins involved, which, at certain times of abnormal gastro-enteric conditions, can get into the circulation in small quan- tity. It is not impossible, furthermore, that such unfortunate cases as the severe forms of angioneurotic edema, which seem, at least in part, to be associated with gastro-intestinal disturbance, and which may be transmitted from mother to child, have their root in anaphy- laxis. For this, however, we have only inference based, on clinical observation. ASTHMA AND HAY FEVER Conditions in which there seems to be more definite ground for association with anaphylaxis are ASTHMA and HAY FEVER. In asthma the analogy has been clearly set forth by Meltzer.22 He points out that in both asthma and anaphylaxis the symptoms consist in a tonic stenosis of the small bronchioli of peripheral origin, and that both conditions are favorably affected by the administration of atropin. It is, of course, not certain, but it seems extremely likely that so-called "nervous asthma" is nothing else than an anaphylactic attack in a hypersusceptible individual when the particular protein to which he is sensitive gains access either by the alimentary or respiratory path. Very closely related to asthma is the condition known as "hay fever." This disease has been of recent years most thoroughly stud- ied by Dunbar.23 Dunbar has ascertained that the hay fever preva- lent in Europe is dependent chiefly upon a protein substance found in the pollen of most grasses, while that of America, which occurs chiefly in the autumn, is caused by the proteins of the pollen cells of the ambrosiacese and solidaginese — plants which are generally dis- tributed on the North American continent and bloom in August and September. The disease occurring in China is caused by another plant, the Ligustrum vulgare. The suggestion that the disease was due to anaphylactic action of these pollen proteins upon hypersus- ceptible individuals was first made by Wolff-Eisner.24 Dunbar has gone into the question with great thoroughness, and has come to the conclusion that the disease has much in common with anaphylaxis — though he believes that, in addition to a hypersusceptibility to the pollen "toxin," there must be present in the patients, at the same 22 Meltzer. Jour, of the A. M. A., Vol. 55, 1910, p. 1021. 23 Dunbar. Berl. klin. Woch., Nos. 26, 28, 30, 1905 ; Zeitschr. f. Immuni- tatsforsch., Vol. 7, 1907; Deutsche med. Woch., Vol. 37, 1911, p. 578. -4 Wolff- Eisner. "Das Heufieber sein wesen u. seine Behandlung," 1906. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 435 time, an abnormal "Durchlassigkeit" or penetrability of the cutis and mucosa for the pollen substances. He claims to have shown that a solution of pollen protein instilled into the eye — or even dropped upon the skin of a hay-fever patient — gives rise to a prompt and severe reaction, while it produces no effect upon normal persons. Unlike experimental serum anaphylaxis, the repeated instillation of the pollen substances rather increases than diminishes the suscepti- bility even when these are carried out daily. Furthermore, unlike serum anaphylaxis, against the manifestations of which no direct passive immunization has so far been possible, Dunbar claims to have produced a curative immune serum by the treatment of horses with the pollen extracts ("Pollantin"). Dunbar, therefore, while admitting an anaphylaxis-like hypersusceptibility of the patients, still believes that the antigen in this case is a true "toxin" against which an antitoxin can be produced — the condition being more directly comparable to the sensitization against diphtheria and tetanus toxins observed during the earlier phases of these investiga- tions by v. Behring and his associates rather than to the phenomena of serum anaphylaxis themselves. Schittenhelm and Weichhardt,25 on the other hand, regard hay fever as truly anaphylactic in every sense. They speak of it as epithelial anaphylaxis (hay fever being specifically designated as "conjunctivitis and rhinitis anaphylactica," in distinction from other forms of cellular anaphylaxis, i. e., enteritis anaphylactica). They believe that the manifestations of the disease result from a local hypersusceptibility in which a toxic substance (Abbau Produkt — similar to anaphylatoxin) is produced. The so-called "antitoxin" of Dunbar acts favorably only when locally applied, and not on sub- cutaneous administration. For this reason they do not regard it as a true antitoxin, but think it acts as a local antiferment which pre- vents or delays the cleavage of the pollen-substance into its toxic split-product — thereby preventing or ameliorating the attacks. Similar to hay fever are the sudden attacks of catarrhal naso- pharyngitis and conjunctivitis — often of asthma-like respiratory dif- ficulty, with itching of the nose and eyes and sneezing which many individuals experience when coming close to horses, cats, or other animals. In the Stanford University laboratory the writer had an assistant who invariably had such attacks, sudden, violent, and of several hours' duration, when handling guinea pigs for experiment. The character of such attacks has long aroused the suspicion that the reaction was anaphylactic in nature, especially since it was known that extremely slight amounts of antigen could give rise to symptoms in susceptible subjects. The difficulty in these cases was the ques- tion of the nature of the antigen which emanated from the animal to excite an attack. Recently, however, observations having impor- 25 Schittenhelm and Weichhardt. Deutsche med. Woch., 37, No. 19, 1911. 436 INFECTION AND RESISTANCE tant bearing upon this problem have been made by Weichhardt 26 and by Rosenau,27 who have demonstrated the presence of organic matter in expired breath. Rosenau condensed the moisture of the expired breath of man and injected the liquid so obtained into guinea pigs. After two weeks these animals were injected with nor- mal human serum, and out of 99 test animals 26 responded with symptoms of anaphylaxis. This demonstrated not only the presence of organic matter in the breath, but showed at the same time that such organic matter was probably protein in nature or at least surely capable of acting as anaphylactic antigen. Rosenau surmises, there- fore, that such protein may be slightly volatile under the given con- ditions. He suggests that sensitization in this manner may explain the harmful effects resulting from a first injection of horse serum into patients, previous sensitization having occurred by close associa- tion with horses. Surely it would explain logically the "cellular" or epithelial anaphylaxis experienced by certain people in the pres- ence of animals. In our opinion this is rendered more likely, since, in the case mentioned as occurring at Stanford University, the in- halation of washings (both aqueous and alcohol soluble) obtained from the hair and skin of guinea pigs, and dried in Petri dishes, produced absolutely no effects in the susceptible individual, whereas continued handling of a living pig almost invariably caused such marked effects that the person in question often became useless as an assistant because of violent attacks of sneezing. It must not be omitted, however, that not all observers have confirmed Rosenau's work, and his explanation must therefore be regarded as merely tenta- tive. An interesting train of suggestions connecting human pathology with anaphylaxis has followed the discovery of "organ-specificity" in the case of hypersusceptibility similar to that noted by TJhlenhuth in connection with the precipitin formation and described in another chapter. It was shown by Kraus, Doerr, and Sohma,28 we have seen, that animals sensitized with the crystalline lens protein of one animal species would react to lens protein in general, and not necessarily to the tissue protein of the animal species from which it was taken. In other words, the ordinary "species" specificity did not hold good. Specificity was determined in this case by the character of the organ rather than by that of the species. The same thing was shown for testicular protein by v. Dungern and Hirschfeld.29 The proteins of these organs from various animals have therefore a certain common antigenic property which is independent of the antigenic element 26 Weichhardt. Arch. f. Hyg., Vol. 74, 1911. 27 Rosenau and Amoss. Jour. Med. Res., Vol. 25, Sept., 1911. 28 Kraus, Doerr, and Sohma. Wien. klin. Woch., Vol. 21, 1908, p. 1084. 29 Von Dungern u. Hirschfeld. Zeitschr. f. Immunitatsforsch., 4, 1910. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 437 common to the particular species. Further than this, Andre jew 30 claims to have shown that it is possible to sensitize an animal with its own lens protein. A few guinea pigs injected by him with their own lens proteins, and reinjected with the same substances after a suitable interval, reacted with definite anaphy lactic symptoms. The possibility is thus given than an animal or human being could become sensitized by its own organ proteins if these were traumatically or otherwise destroyed and absorbed. The train of reasoning is similar to that which has given much hope of enlightenment to pathologists when the earlier work upon the cytotoxins was done. Rosenau and Anderson,31 for instance, injected guinea pigs with guinea pig pla- centa, and found that, after the usual period of incubation, the ani- mals reacted to a second injection with marked symptoms of anaphy- laxis. On the basis of these experiments Rosenau and Anderson suggest that certain of the toxemias of pregnancy are of anaphylactic origin. They believe that it is possible that a mother may become sensitized by the "autolytic products of her own placenta/7 the result being eclampsia. By a similar process of reasoning Elschnig32 has attempted to explain sympathetic ophthalmia. He claims to have shown that the laws of organ specificity apply to the proteins (especially the pig- ment) of the uveal tract. The destruction and absorption of injured uveal tissue, according to him, induce the formation of organ- specific antibodies by which the remaining uveal structures of the same, as well as of the opposite, eye are sensitized. The consequence is a "sympathetic" inflammation which "is to be regarded purely as an anaphylactic reaction." These and other similar suggestions less well founded experi- mentally illustrate the possibilities for clinical reasoning furnished by a knowledge of the anaphylactic phenomena. In no cases of this sort, however, can the association with anaphylaxis be as yet re- garded as more than an extremely interesting suggestion. From all that has gone before it is quite evident that most of the positive facts which may be regarded as determined concerning the phenomena of anaphylaxis have been obtained in experiments with small and very sensitive animals, comparatively large and measured quantities of antigen, and often by the violent method of intravenous injection in which the entire mass of antigen comes rapidly into contact with the available antibodies and the vulnerable tissues. We cannot, therefore, draw rigid parallels between these experiments and clinical manifestations in human beings in whom 30Andrejew. Arb. a. d. kais. Gesundh., Vol. 30, 1909. 31 Rosenau and Anderson. U. S. Pub. Health and M. H. S. Hug. Lab. Bull 45, 1908. 32 Elschnig. Von Graefe's Archiv f. OphthaL, Vol. 75, p. 459 ; Vol. 76, p. 509; Vol. 78, p. 549. 438 INFECTION AND RESISTANCE the localization, quantitative discharge of antigen, and consequent production of antibodies is of necessity irregular and different in each individual case. We have learned, as a general conception, however, that the introduction into the animal body of antigenic sub- stances of all varieties leads, under certain conditions, to increased tolerance or resistance — under other circumstances to a state of greater susceptibility — these diametrically opposed physiological consequences being, in all probability, determined by relative concen- trations of antigen and antibody, their speed of contact, and their quantitative relationship to available alexin. The problems of clini- cal medicine in which the possibility of anaphylaxis can be at all considered, therefore, are extremely complicated, and few of them can be approached by direct experiment. In the case of serum sickness the analogy has been so clear and the experience with human beings so extensive that practically no doubt can exist as to the common mechanism of this condition with that of experimental anaphylaxis. In the other conditions men- tioned the connection is one of great likelihood, but after all is in- ferential, and calls for much further investigation. For this reason it is best to abstain from a further enumeration of many other maladies in which the condition of hypersusceptibility has been sug- gested as a vaguely possible etiological factor. ANAPHYLAXIS AND THE TUBERCULIN REACTION There is one class of phenomena, however, which calls for further discussion in this connection, since its dependence upon anaphylaxis, while generally assumed, is still opposed by many authorities. This consists of the various DIAGNOSTIC REACTIONS in which extracts of micro-organisms are injected, or brought into con- tact with the skin or conjunctiva of infected subjects. Such are the various forms of the tuberculin reaction, the typhoid reaction of Chantemesse, the one of Gay, and the luetin reaction of Noguchi. In. the tuberculin reaction the conditions have been thoroughly studied, and we may make a detailed consideration of this example serve to bring out the general principles involved. In all forms of the TUBERCULIN REACTION there is a very evident hypersusceptibility to various forms of antigen derived from the bacillus. When the tuberculin is injected subcutaneously the reac- tion is systemic and also localized to a certain extent in any tubercu- lous foci which may be present. When the v. Pirquet or Moro skin reactions are carried out, or the Calmette ophthalmic test is made, the reactions are almost purely local. In all cases reactions are induced by quantities of antigen which cause no effect whatever in normal individuals. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 439 The -basic observation leading to the diagnostic use of tuberculin was made by Koch 33 upon guinea pigs. He describes his observa- tion as follows: "Tuberculin may be injected into normal guinea pigs in consid- erable quantities without causing noticeable symptoms. Tuberculous guinea pigs, on the other hand, react to comparatively small doses in a very characteristic manner." Since, in Koch's experiments upon tuberculin, it was desirable for his particular purposes at the time, to obtain very sharp reac- tions, he did not content himself with the production of moderate symptoms by the injection of slight amounts of tuberculin into in- fected animals, but increased his dosage until the guinea pigs were killed. He showed that guinea pigs having a moderately advanced infection — 4 to 5 weeks after inoculation — could be killed by doses of 0.2 to 0.3 gram, while animals in very advanced stages would suc- cumb within 6 to 30 hours to quantities as small as 0.1 gram sub- cutaneously. In the animals so studied he determined not only a systemic effect, but a very marked local reaction as well in the skin, areolar tissues, and adjacent lymph nodes. Koch's observations upon guinea pigs were applied by him, Gutt- stadt,34 Beck,35 and others to man, and the result was the develop- ment of the present important diagnostic test. The fundamental fact in this as well as in other tests of this kind, then, is the appearance of local and systemic reactions in infected subjects to con- tact with specific antigenic material which, at least in the same quantities, produces no effects in normal individuals. The analogy with the phenomena of anaphylaxis is thus indicated. Koch's original interpretation of the phenomenon was of course unaided by any of the later observations on anaphylaxis. According to him the tuberculin contained substances which caused tissue necrosis. The necrotizing action was particularly powerfiil upon tissues which were tuberculous, and therefore already saturated with the toxic material. The destruction of such tissues resulted in sys- temic symptoms. Very similar to this view is the one later expressed by Babes and Broca,36 who attribute the systemic symptoms to a sudden lighting up of the existing lesions by the small amount of extra tuberculin added to that already present in these foci. The first suggestion of the possible association of the tuberculin reaction with the union of an antigen and its antibody was made by Wassermann and Bruck.37 They accepted Ehrlich's assumption 33 Koch. Deutsche med. Woch., No. 43, 1891. 34Guttstadt. "Klin. Jahrbuch" Erganzungsband, 1891. 35 Beck. Deutsche med. Woch., No. 9, 1899. 36 Babes u. Broca. Zeitschr. f. Hyg., Vol. 23, 1896. 37 Wassermann and Bruck. Deutsche med. Woch., No. 12, 1906. 440 INFECTION AND RESISTANCE that certain cells of the tuberculous foci (those situated just below the periphery and already affected by the tubercle toxin, though still resistant) were possessed of an increased receptor apparatus for the tubercle antigen. For this reason the injected tuberculin was con- centrated in these foci, attracted out of the circulation by the in- creased avidity of these cells, the consequence being increased ac- tivity of the lesions and systemic symptoms. The tuberculin reac- tion, according to these writers, therefore, would be caused by the union of the tuberculin with the "sessile receptors" upon the diseased tissues — a point of view which would specify the diseased tissues and their products as the sources from which emanated the toxic factors inciting the systemic symptoms. The theories of Koch and of Babes do not, as Meyer points out, explain the frequent absence of the tuberculin reaction in very ad- vanced cases of human tuberculosis, as contrasted with its frequency and regularity in the earlier cases. For, according to both of these views, the more severe the existing lesions the more actively would the injected tuberculin initiate tissue necrosis and consequent symp- toms. The theory of Wassermann and Bruck avoids this objection since it presupposes the acceptance of Ehrlich's view that the in- creased receptor apparatus is present and free only in those cells in which necrotic destruction has not yet set in. In the necrotic areas the receptor apparatus is already saturated or satisfied as to its affinities, and extensive areas of necrosis, therefore, are unaffected by contact with further quantities of tuberculin. All of these theories, however, inasmuch as they refer the tubercu- lin reaction to alterations taking place in more or less active lesions, are unable to account for the occurrence of the reaction in persons in whom healed foci only are present, and are entirely inconsistent with the facts we now possess regarding the cutaneous and ophthal- mic tests in which the reactions occur in previously healthy tissues. These facts practically exclude the acceptation of any theories which regard the tuberculous focus as the sole source of the reaction. We may still accept the Koch or Wassermann views to explain local swellings and other changes in infected lymph-nodes or other lesions, but we must assume in addition to this a generalized hypersuscepti- bility at least analogous to the phenomena of anaphylaxis. This ability of previously healthy tissues, remote from any center of tuberculous infection, to react to the application of tuberculin was discovered by von Pirquet38 in the development of his skin reaction, and by Calmette 39 and Wolff-Eisner 40 in their work upon 38 v. Pirquet. Berl klin. Woch., No. 20, p. 644, and No. 22, p. 699, 1907; also "Klinische Studien iiber Vaccination," Deuticke, Wien, 1907. 39 Calmette. C. R. de I'Acad. des Sciences, June, 1907. 40 Wolff-Eisner. Berl. klin. Woch., 1907, p. 1052. Discussion of paper by Citron. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS the ophthalmoreaction. The principle involved in these reactions was then further emphasized by the introduction of the Moro tuber- culin-ointment method and the intracutaneous test of Romec. The mere observation that the infection with tuberculosis results in a general tissue hypersusceptibility immediately suggests the interpre- tation of the tuberculin reaction as a manifestation of anaphylaxis. Von Pirquet, accordingly, on the basis of his previous studies upon serum sickness includes the reaction in the category of what he calls, "allergic." He assumes that the reaction depends upon the presence in the system of antibodies, which form a union with the applied tubercu- lin, the result being the formation of poisons and a reaction. This assumption, according to the anaphylatoxin theory of Friedberger, would imply the participation of alexin in the reaction — acting upon the united tuberculin-antituberculin complex, though v. Pirquet does not express himself positive as to this. This, moreover, is the clearly expressed opinion of Friedberger 41 himself. Consistently with his general theory of anaphylaxis he assumes that the injected tuberculin comes into relation with specific antibodies with which it unites, the alexin then splitting off anaphy- latoxin from the complex. He bases this view upon his experimental demonstration, mentioned above, of the production of anaphylatoxin from tubercle bacilli by in vitro digestion with guinea pig comple- ment. In principle the view of v. Pirquet is similar to that previously expressed by Wolff-Eisner42 that the union of tuberculin with its lytic antibody, present in the tuberculous animal, gave rise to pois- ons as the result of lysis. Both of these theories simply apply to the special case of the tuberculin reaction theories of mechanism applied to anaphylactic reactions in general. We must admit that the facts of the "allergie" reactions as a class seem to force upon us the acceptation of von Pirquet's views. Apart from the purely clinical observations made in carrying out the routine tests we have the additional evidence that the instillation of tuberculin into the eye of normal individuals gives rise to no reac- tion, but a repetition of the instillation into the same eye after ten or more days results in a marked and typically positive test. Further- more, von Pirquet 43 states that individuals showing no clinical tu- berculosis and negative to a first test will often react ("sekundare Reaktion") to a second test carried out a few days after the first. These facts all seem to indicate acquired hypersusceptibility more analogous to true serum-anaphylaxis than to the toxin hypersuscepti- 41 Friedberger. Munch, med. Woch., Nos. 50 and 51, 1910. 42 Wolff- Eisner. Berl klin. Woch., Nos. 42 and 44, 1904. 43 Cited from Lowenstein in "Kraus u. Levaditi Handbuch," Vol. 1, p. 1039. 442 INFECTION AND RESISTANCE bility of von Behring, in that the tuberculin is but slightly toxic in itself. If the analogy is such a close one, therefore, it should be easy to formulate experiments by which the phenomena now ascertained regarding serum-anaphylaxis could be demonstrated for tuberculin hypersusceptibility. The obvious procedure, therefore, would be to attempt to passively transfer tuberculin sensitiveness to a normal animal with the serum of a tuberculous one. This has indeed been attempted by Friedemann,44 later by Bauer 45 and a number of others — usually with negative result. The writer, hoping to develop a diagnostic method for tuberculosis, has also attempted this by the transference of human tuberculous blood to guinea pigs, but invari- ably obtained negative results. Yamanouchi 46 alone has reported positive experiments by a similar procedure with rabbits, but so far his results, according to Friedemann, have completely failed of con- firmation. Austrian succeeded by sensitizing guinea pigs with 5 c. c. of titrated whole blood, using for the second injection a tuber- culo-protein prepared by the method of Baldwin.47 In this particu- lar, therefore, the analogy between anaphylaxis and the tuberculin reaction, though not easily worked out, has nevertheless been estab- lished. Another objection which has been made by a number of ob- servers is the fact that anaphylaxis is accompanied by temperature depression while tuberculin reactions are followed by a rise. This objection may be regarded as invalid, however, in the light of Fried- berger's 48 experiments which showed that temperature depression follows only when large doses of the antigen are injected into the sensitized animals, smaller doses often giving rise to increased tem- perature. We gain a certain amount of insight into the conditions here prevailing by considering the information which has been obtained from the study of the antibodies formed in animals in tuberculosis. It appears from the work of Christian and Rosenblatt 49 that anti- bodies to the tubercle bacillus are formed by tuberculous animals only. Normal animals form these to a very slight degree only, if at all, when immunization with tuberculin is attempted. In other words, as Friedemann ("Weichhardt's Jahresbericht," 6, 1910) points out, the specific reaction of antibody formation in tuberculosis seems to be closely associated with the tuberculous tissues themselves. 44 Friedemann. "Uber anaphylaxie," "Weichhardt's Jahresbericht/' Vol. 6, 1910. 45 Bauer. Cited ibid.; also Munch, med. Woch., 1909, p. 1218. 46 Yamanouchi. Wien. klin. Woch., 1908, p. 1263. 47 Austrian Bull of the Johns Hop. Hosp., Vol. 24, 1913 ; Baldwin, Journ. Med. Res., Vol. 17, 1910. 48 Friedberger. Deutsche med. Woch., No. 11, 1911. 49 Christian and Rosenblatt. Munch, med. Woch., 1908. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 443 The same inference can be made from Bail's 50 experiments on passive sensitization. For, although passive sensitization of guinea pigs with the serum of tuberculous animals has been unsuccessful, Bail succeeded in obtaining lethal anaphylactic reactions by injected macerated tuberculous tissues, following these on the next day by injections of tuberculin. It is plain from this, as Friedemann cor- rectly argues, that we must assume that the antibodies (receptors) formed against tubercle bacilli are closely bound up with the tissue cells, the reaction of tuberculin being largely with " sessile receptors." Indeed, it seems as though the antibodies formed against tubercle bacilli undoubtedly remain in close relation to the cells of the tis- sues except in cases of active tuberculosis in which localized areas of cells are under the influence of a very intense action of the poi- sons, and a consequent overproduction and discharge of receptors (using the Ehrlich nomenclature) may occur. This would corre- spond with considerable accuracy, moreover, to the histological facts, for in this infection, similar to leprosy and a number of other con- ditions, but unlike most acute infections, the battle against the micro- organisms is carried out chiefly by the adjacent tissue cells. We might assume, therefore, that, in tuberculous individuals, there is indeed a reaction, at first local, then generalized to a slight degree, in which antibodies are actually formed. These antibodies, however, remain to a preponderant extent sessile, or incorporated in the reacting cells. Upon the injection or application of tuberculin the reaction takes place in or upon the cells. Whether or not the further cooperation of complement or alexin is then necessary for the lysis and poison production from the antigen, as in the similar reac- tions taking place in the circulation, or whether the intracellular ferments themselves suffice for this, cannot be decided at present. It is certainly not unlikely that the circulation of tubercle-antigen— even in small quantities — throughout the body may produce such hypersusceptibility of cells (represented graphically by the concep- tion of sessile receptors in many parts of the body remote from the lesion — a quality remaining constant for prolonged periods and ex- plaining the subsequent skin and ophthalmic reactions obtained upon test). Certain clinical observations cited by v. Pirquet 51 would seem to support such a view. For instance, he states that, having em- ployed his left forearm repeatedly for tests, he was able to obtain positive reactions in this area with tuberculin diluted 1 to 1,000, whereas his right forearm was negative to tuberculin ten times as concentrated. Furthermore, as Kohn 52 has shown that, while the first injection of tuberculin into the eye of a normal person produces no reaction, this eye will not only react to a second instillation, 50 Bail. Zeitschr. f. Immunitatsforsch., Vol. 4. 11 V. Pirquet, "Kraus u. Levaditi Handbuch," Vol. 1, p. 1050. 52 Kohn. Quoted from Lowenstein, Kraus and Levaditi, Vol. 1, p. 1033. 444 INFECTION AND RESISTANCE but will show a reaction when the second application is made sub- cutaneously. Negative evidence pointing in the same direction is the observation that absolutely no influence is exerted upon the out- come of tuberculin reaction if the tuberculin is previously mixed with blood serum from either positively or negatively reacting cases.53 This would tend to show that, whether reacting or not, the factors which determined this are certainly not present in the cir- culating plasma. That the circulation of tubercle antibodies in the blood may even interfere with the localized tuberculin reaction is rendered likely by the fact that skin reactions are often negative in cases of advanced tuberculosis, and that, as we are told by Dr. Blair, of the N. Y. Zoological Park, such reactions are usually negative in tuberculous monkeys in which the disease is invariably very rapid and acute. PRACTICAL DIAGNOSTIC USES OF ANAPHYLAXIS *The specificity of the anaphylactic reaction has led to extensive attempt to utilize it for forensic protein determinations in the same way in which the precipitin test is used. Uhlenhuth,54 Thomsen,55 Pfeiffer, and others have carried on extensive experimentation in this problem, the technique, in general, consisting in sensitizing guinea pigs with solutions of the unknown protein (dissolved blood spots, etc.) and testing them by a second injection of the suspected protein after the usual anaphylactic incubation time. The results of such work have shown that indeed positive reaction may be obtained and diagnosis made in this way. However, the reactions are not ordi- narily very striking, and this method is not as reliable as the precipi- tin method. Uhlenhuth 56 believes that the anaphylactic reaction has value only in cases in which the amount of unknown protein is so small or so changed by preservation or decomposition that its pre- cipitable qualities have been lost. Yamanouchi's 57 attempt to utilize anaphylaxis for the diagnosis of tuberculosis, by passively sensitizing guinea pigs with the serum of tuberculous patients and testing subsequently with tuberculin, has been mentioned before. Although he claims positive experiments, our own experience with a similar technique has given us results which were so irregular that we feel that this technique has very slight practical value, if any. Pfeiffer 58 has attempted to apply anaphylaxis to the diagnosis 53 V. Pirquet. Loc. cit. 54 Uhlenhuth. Deutsche milit. Zeitschr. No. 2, 1909. Cited from same author, Zeitschr. f. Immunit'dtsforsch., Vol. 1, 1909. 55 Thomsen. Zeitschr. f. Immunitatsforsch., Vol. 1, 1909. 56 Uhlenhuth. Zeitschr. f. Immunitatsforsch., Ref. Vol. 1, 1909, p. 525. 57 Yamanouchi. Wien. klin. Woch., No. 47, 1908. 58 Pfeiffer. Zeitschr. f. Imm., Vol. 4. 1910. CLINICAL SIGNIFICANCE OF ANAPHYLAXIS 445 of malignant tumors. Together with Finsterer 59 he sensitized guinea pigs with the serum of carcinomatous patients following the injection 48 hours later with press-juices of tumors. His conclu- sions were drawn from the anaphylactic temperature reaction, and he claims that animals so sensitized are hypersusceptible to the juices obtained from carcinomata, whereas animals sensitized with normal serum or the serum of sarcoma patients show no hypersus- ceptibility. Ranzi 60 has not been able to confirm this. The signifi- cance of such experiments, if correct, apart from their practical value for the diagnosis of carcinoma, would be considerable in that they tend to show that cancer tissues contain a specific protein which is antigenically distinct from the other tissue-proteins of the afflicted individual. However, we cannot yet accept these facts as absolutely established. 59 Pfeiffer and Finsterer. Wien. klin. Woch., No. 28, 1909. 60 Ranzi. Zeitschr. f. Imm., Yol. 2, 1909. CHAPTER XIX THEKAPEUTIC IMMUNIZATION IN MAN FACTS CONCERNING ANTITOXIN TREATMENT IN. MAN THERAPEUTIC USE OF DIPHTHERIA ANTITOXIN IT is not consistent with the purpose of this brief treatise to dis- cuss extensively the therapeutic benefits obtained by serum therapy in diphtheria. We can convey briefly an adequate idea of this by citing some of the tables given by Northrup in Nothnagel's "Encyclopedia of Practical Medicine/' American Edition, Volume on Diphtheria, etc., p. 143. These figures are taken from the statistics of the New York Board of Health, which began treatment of diphtheria with antitoxin in January, 1895. Dr. Northrup states, however, that serum treatment cannot be considered to have been in general use until some time later. Without Antitoxin Year Cases reported Deaths Mortality, per cent. 1891 5,364 1,970 36 7 1892 5,184 2,106 40 0 1893 . . . 7,021 2,558 36 4 1894 9,641 2,870 29 7 Total 27,210 9,504 Avg. 34 9 With Antitoxin 1895 10,353 1,976 19 0 1896 11,399 1,763 15 5 1897 10,896 1 590 14 5 1898 7,173 919 12 8 1899 8,240 1,085 13.1 1900 8,364 1,176 14 0 Total 56,425 8,509 Avg. 15 0 Table taken directly from Northrup, loc. cit. 446 THERAPEUTIC IMMUNIZATION IN MAN 447 From this table there appears a reduction of 58 per cent, in mortality and a similar drop is evident from the German statistics of Dieudonne,1 from those of Welch, and many others. It should be considered, moreover, in reading such statistics that they are made on gross mortality reports without elimination of the many cases that have not come under observation until too severely diseased to react to any form of treatment. The reason for the fail- ure to obtain results with antitoxin when the cases have proceeded beyond a certain stage of intoxication will become evident when we consider the manner of absorption of the poison in a succeeding para- graph. The mortality sinks to between 8 and 9 per cent., when such cases are omitted, as is shown by the collective investigations of the American Pediatric Society in 1896 — figures which we take also from i^orthrup's comprehensive study. This purely statistical evi- dence, however good, is further reenforced by the unquestionable and considerable diminution of emergency operations,2 such as intu- bation and tracheotomy, since introduction of the antitoxin. More- over, there is the manifold clinical evidence of benefit, after the serum treatment, familiar to every practicing physician. Although the injection of antitoxin is of benefit by whatever route and in whatever quantity it may be given, nevertheless recent experimental investigations have taught us much regarding the proper use of this therapeutic agent. Especially interesting are the investigations of Meyer,3 who showed the extreme importance of an early use of the antitoxin. Apparently, as we have mentioned in another place, like tetanus antitoxin, the diphtheria poison may be in part absorbed directly by the nerves.4 There is apparently a great difference in therapeutic efficiency, according to the method in which the serum is administered, a differ- ence probably depending upon speed of absorption. Berghaus5 showed that intravenous injection is 500 times more potent therapeu- tically than the subcutaneous, and 80 to 90 times more so than the intraperitoneal injection. Schick, in discussing this problem from the clinical point of view, for this reason lays special stress upon the speed of administration. He says: "Not only days but hours are of great importance." He bases this opinion largely upon the fact that the toxin which has already united with the nerve substance can probably no longer be influenced by the injection of the serum. According to the experiments of Meyer and Hanson diphtheritic 1 Dieudonne". Arb. aus dem kais. Gesund., XIII, 1897. 2 Siegert. "Jahrbuch f. Kinderheilkunde," Vol. 52, cited after Wernicke. 3 Meyer. Berl. Tel. Woch., 25, 26, 1909 ; Arch. f. exp. Path. u. Ther., Vol. 60, 1909, and Berl. kl. Woch., No. 45, 1911. 4 For a thorough discussion of these conditions see Schiek, Centralbl. f. Bakt., Rev. Vol. 57, 1913, "Report of 7th Meeting of the Mikrobiol. GeselL," Berlin, 1913. 5 Berghaus. Cited from Schick, loc. cit. 448 INFECTION AND RESISTANCE paralysis may follow even when vigorous serum treatment has been employed. For, according to them, only the toxin which has reached the central nervous system through the circulation can be influenced by the serum, but no effect is possible upon the fraction which has been absorbed from the nerve endings directly. Schick,6 on the basis of extensive experiments, comes to the con- clusion that the subcutaneous injection of 1,000 to 2,000 units in diphtheritic cases has an immunizing value only, which protects the tissues from further injury and leads to cure if, at the time of injec- tion, the lethal dose has not yet united with the sensitive cells. "If," he states, awe wish to obtain antitoxic action upon toxin which has already gone into action before the injection of the serum, then re- sults can be obtained both in man and in animals only if a great deal of antitoxin is injected intramuscularly or intravenously." 7 Interesting also from a clinical point of view are the studies of Schick,8 Hahn,9 and others10 upon the presence of antitoxin in the blood of normal, untreated individuals at different ages. These investigations were carried out by the intracutaneous method of toxin and antitoxin determination described in greater detail in a later section. The following table, taken from the article of Hahn, illustrates the experience, in such investigations, both of Schick and of Hahn himself. The determinations were carried out upon indi- viduals who had never had diphtheria, as far as could be learned. Age Cases with antitoxin serum Cases without antitoxin serum Highest antitoxin value in 1 c. c.u Schick - r Newborn .... ^0-1 year 11 1 0 3 under 1.5 units 0.11 unit r 2-10 years.. 7 5 1.0 unit 11-20 years.. 8 9 0.75 unit Hahn 21-30 years 9 5 2 . 5 units 31-40 years. . 5 1 0.25 unit 1 41-65 years . . 2 8 2 . 5 units The table shows that in newborn children there is almost regu- larly a definite and sufficient protective value in the serum which diminishes up to the first year, so that at the end of the first year three out of four individuals had no antitoxin in their serum. In subsequent years up to the age of 40 an increasing percentage of 6 Schick. Loc. cit. 7 Schick. Loc. cit., p. 32. 8 Schick. "tiber Diphtherimmunitat," Wiesbaden, 1910. 9 Hahn. Deutsche med. Woch., Vol. 38, No. 29, p. 1366, 1912. 10 Karasawa and Schick. Zeitschr. f. Kinderkranklieiten, 1910, and "Jahr- buch f. Kinderheilkunde," 1910. 11 Table taken directly from Hahn, loc. cit. THERAPEUTIC IMMUNIZATION IN MAN 449 people have sufficient amounts of diphtheria antitoxin in their blood. After the age of 40 an increasing percentage is without such protec- tion. The first observation, that newborn children usually possess considerable amounts of antitoxin, is very probably due to passive immunization by the blood of the mother, a fact which we have men- tioned in another place. The exact method by which such measure- ments are made is described in a subsequent section on the intra- cutaneous method of determining toxin and antitoxin. The work of Schick, that of J. Henderson Smith, and recent studies by Park and Biggs promise to alter considerably the methods of antitoxin therapy as at present in use in diphtheria. Smith meas- ured the speed of absorption of antitoxin injected subcutaneously into the abdominal wall of a healthy man. His results are shown in the following table, which we take from his communication (page 213) : TABLE V One c. c. of the patient's serum contained: Before injection No demonstrable antitoxin 5 hours after injection 0.1 unit antitoxin 14 hours after injection 0.225 unit antitoxin 32 hours after injection 0.68 unit antitoxin 44 hours after injection 1.0 unit antitoxin 3 days after injection 1.3 units antitoxin 4 days after injection 1.3 units antitoxin 6 days after injection 0 . 68 unit antitoxin 13 days after injection 0 . 17 unit antitoxin 15 days after injection 0 . 14 unit antitoxin 20 days after injection 0 . 08 unit antitoxin 27 days after injection No demonstrable antitoxin12 Park and Biggs 13 have made similar studies and have contrasted the speed of absorption after subcutaneous administration with that after intravenous injection, basing their curves upon careful measure- ments of the sera of the treated patients. We reproduce their charts as given in their recent publication. It is apparent from these charts, as well as from the work of Hen- derson Smith, that antitoxin, subcutaneously given, is slowly ab- sorbed, and does not reach its maximum concentration in the blood stream until forty-eight hours or more after the injection. It fol- lows that, as Park and Biggs point out, it is more rational to inject a single adequate dose than to divide the dosage and inject at inter- vals. They have obtained results in animal experiment which graphically illustrate this principle. A rabbit which had received ten fatal doses of toxin intravenously was given a total of 500 anti- toxin units in divided doses as follows : 100 units after twenty min- 12 J. Henderson Smith. Journal of Hygiene, Vol. 7, 1907, p. 205. 13 Park and Biggs. Collected Studies from the N. Y. Department of Health, Bureau of Laboratories, Vol. 7, 1912-1913, p. 27. 450 INFECTION AND RESISTANCE I 2 CNIT9 CHART I. — Showing the extent and rapidity of absorption of 10,000 units of antitoxin given subcutaneously. Each line represents the antitoxin content of 1 c. c. of blood at different intervals of time. (From Park and Biggs,, loc. cit.) THERAPEUTIC IMMUNIZATION IN MAN 451 UNITS CHART II. — The antitoxic power of human blood after an intravenous injection of 10,000 antitoxic units. (From Park and Biggs, loc. cit.) utes, 100 after 40 minutes, and 150 units each after 60 and 80 min- utes. This rabbit died. Another animal given the same dose of toxin received 200 units of antitoxin twenty minutes later and lived. The amount necessary to save life in rabbits receiving ten fatal doses intravenously was as follows : Given after 10 minutes 5 units antitoxin Given after 20 minutes 200 units antitoxin Given after 30 minutes 2,000 units antitoxin Given after 45 minutes 4,000 units antitoxin Given after 60 minutes 5,000 units antitoxin Given after 90 minutes . . No amount These extremely important experiments of Park and Biggs bear out the opinion of Schick and show beyond question that the proper way to give antitoxin is to give a single adequate dose as early as possible. They emphasize the fact that probably the most important single point in the specific therapy of diphtheria is the speed with which the diagnosis can be made and the antitoxin given. At the De- partment of Health the dosage now employed, as given by Park and Biggs, is the following : UNITS IN CASES Mild Moderate Severe Very Severe Infants under 1 year Children 1 to 5 years 2,000 3,000 3,000 5,000 10,000 10,000 10,000 10000 Children 5 to 9 years Persons over 10 years 4,000 5,000 5,000 10,000 10,000 10,000 15,000 20,000 452 INFECTION AND RESISTANCE PRACTICAL CONSIDERATIONS CONNECTED WITH DIPHTHERIA ANTI- TOXIN PRODUCTION AND STANDARDIZATION The conditions which govern the active production of toxins by bacteria in culture media are not only of great theoretical interest hut possess unusual practical value in that the most important factor for the successful production of a strong antitoxin consists in the preliminary preparation of a potent toxin. The bacterial true toxins are all "exotoxins" in that they are soluble, moderately diffusible substances which pass readily from the bacterial bodies to the en- vironment, and for this reason can be obtained most readily by the cultivation of the bacteria upon fluid media and subsequent nitration of the cultures through earth or porcelain niters. The choice of culture or strain is an important element in this procedure, since within the same species of toxin-producing micro- organisms there is much variation in the speed and energy of toxin production. Thus for unknown reasons some strains of diphtheria bacilli will far outstrip others in this respect. An excellent illustra- tion of this is the experience of Park and Williams 14 with two diph- theria cultures — a very virulent and a very weak one. Of the former, 0.002 c. c. of a forty-hour bouillon culture killed a guinea pig, while of the latter 0.1 c. c. of a similar culture was necessary for the same result. In the case of tetanus, cultural differences do not seem to be as common. Individual strains also may gain or lose in toxin-pro- ducing powers, according to the method of handling them which is practiced. It is stated,15 for instance, that a diphtheria culture will lose in energy of toxin production if permitted to grow without suffi- ciently frequent transplantation. However, transplanted on solid media with reasonable frequency, these bacteria show a remarkably constant toxin production. A well-known strain, the Park-Williams No. 8, now in use in many antitoxin laboratories throughout the world, has persisted for over 15 years in producing a strong toxin. There are occasional strains among toxin-forming species which are entirely devoid of this property. Diphtheria bacilli which were virulent while possessing all the other cultural characteristics of the group have been described, but appear, from the experience of the writer, to be rather uncommon.16 Of tetanus bacilli little is known in this respect. Given a powerfully toxic strain of the proper bacteria the method of cultivation is also of great importance in influencing the eventual yield of poison. These relations have naturally been stud- ied with the greatest care in the case of diphtheria and tetanus 14 Park and Williams. "Pathogen. Micro-organ.," N. Y., 1910. 15 Park and Williams. Loc. cit. 16 Zinsser. Jour. Med. Bes., N. S., Vol. 12, 1907. THERAPEUTIC IMMUNIZATION IN MAN 453 bacilli, since in these cases there has been the greatest practical appli- cation for such knowledge. In the case of diphtheria, though toxin will be produced on all media on which the bacillus grows easily, the most favorable medium for this purpose is a slightly alkaline broth made of lean beef or veal infusion and containing peptone. Since acid formation hinders the production of toxin, Martin 17 has suggested fermentation of the APPAEATUS ARRANGED FOR THE STERILE FILTRATION OF DIPHTHERIA CULTURES IN TOXIN PRODUCTION. (After Kosenau, U. S. Hyg. Lab. Bull. 21, 1905, p. 38.) muscle sugar with yeast, while Theobald Smith 18 recommends pre- liminary fermentation with Bacillus coll. Park and Williams 19 regard this as unnecessary. They recom- mend a 2 per cent, peptone broth made of veal. This is neutralized to litmus and 7 to 9 c. c. of normal NaOH solution to the liter are added. In such a medium at 37.5° C. the production of toxin begins within 24 hours and reaches its highest point in from five to ten days. When at its height the process must be stopped and the cul- tures exposed to a lower temperature, otherwise rapid deterioration takes place because of the instability of the toxin. Even when kept cold and in the dark this deterioration proceeds steadily though slowly. At first, however, even under these conditions a compara- tively extensive loss of toxin goes on — a process sometimes spoken of as "maturing of the toxin" — after which the poison strikes a 17 Martin. Ann. Past., 1896. 18 Th. Smith. Journ. Exp. Med., IV, 1899, p. 373. 19 Park and Williams. Journ. Exp. Med., Vol. 1, 1896. 454 INFECTION AND RESISTANCE fairly constant and very gradual rate of weakening, and is, com- paratively speaking, stable. In the United States Hygienic Laboratory in Washington, ac- cording to Rosenau, the recommendations of Theobald Smith are largely followed in the production of toxin. The procedure is as follows : The culture medium, "Smith's Bouillon," is prepared from chopped beef from which fat and tendon have been cut out. This is adjusted by phenolphthalein titration to 0.5 per cent, acidity. It is then placed into Fernbach flasks and inoculated on the surface with a Park-Williams bacillus ~No. 8. The flasks are incubated for 7 days at 37.5° C. The reaction of the medium after such incubation is determined, and flasks showing an acidity of 1.5 or over are dis- carded. The usual reaction at the end of incubation is 0.6 to 0.8 per cent, acidity. This broth is filtered through Berkefeld filters or porcelain candles. Toxin so prepared is now tested and its L0 and L+ doses deter- mined by the methods described above. Rosenau20 states that poi- sons are discarded as containing too large a proportion of toxon if the difference between L0 and L+ is greater than 15 M L D. The toxin is now set aside in flasks for the process which Rosenau calls "seasoning." At intervals of about a month it is retested and finally it is found that the rate of toxoid formation decreases and the poison reaches a period of equilibrium. It can now be used for accurate determination of the L+ dose, and this is done from careful measurements on a large number of guinea pigs. Examples 21 of such measurements, abbreviated for the sake of simplicity, are given in the following tables : Toxin Determinations of M L D or "T" Dose in c. c. Result 0.03 = death in 1^ days 0.02 = death in \y% days 0.01 = death in 2 days 0.008 = death in 3 days 0.006 = death in 3^ days 0.005 = death in 4 days M LJ> 0.004 = death in 6 days 0.003 = death in 8 days 0.002 = late paralysis 0.001 = well in 16 days. Toxin Determination of L+ Dose 1 Antitoxin unit + 0.2 c. c. = 0 1 Antitoxin unit -j- 0.21 c. c. = 0 = LO * 1 Antitoxin unit -j- 0.22 c. c. = local infiltration 20 Rosenau. Hyg. Lab. Bull. No. 21, April, 1905. 21 Examples are taken from measurements reported by Rosenau, loc. cit. THERAPEUTIC IMMUNIZATION IN MAN 455 1 Antitoxin unit + 0.23 c. c. = fatal in 17 days 1 Antitoxin unit -j- 0.24 c. c. = fatal in 14 days 1 Antitoxin unit -f 0.26 c. c. = fatal in 9 days 1 Antitoxin unit -j- 0.28 e.c. = fatal in 6 days i Antitoxin unit + 0.29 c. c. = fatal in 4 days = L+ 1 Antitoxin unit + 0 r3~"c. c. = fatal in 3 days The production of antitoxin is carried out by the graded injec- tion of antitoxin into horses. Young, healthy horses are chosen, tested for freedom from glanders, and the first injections are made either with toxin attenuated by the addition of Lugol's solution or terchlorid of iodin, or, as in the New York Health Department, the first injections consist of mixtures of toxin and antitoxin. We take our description largely from the account given by Park.22 The first injection consists of 12 c. c. of toxin (M L D 1/400 c. c.), together with 100 units of antitoxin. After the reaction from such an injec- tion has completely subsided — after 3 to 5 days — a second injection is given of toxin without antitoxin; then 15 c. c., 45 c. c., 55 c. c., 65 c. c., 80 c. c., 95 c. c., 115 c. c., 140 c. c., etc., the intervals be- tween injections being about three days and depending upon the reaction of the horse and the speed with which it entirely recovers from the preceding injection. In a particular case cited by Park 675 c. c. of toxin could be given by the 60th day ; in this case by the 28th day the horse was yielding 225 units to the c. c. ; on the 40th day, 850 units ; on the 60th day, 1,000 units. The determination of the antitoxin unit, carried out from time to time on the serum of such a horse against the L+ dose described in our preceding table, would be carried out as follows : In all such standardization great care must be taken in employ- ing accurately standardized glassware. Rosenau recommends em- ploying "capacity instruments" rather than "outflow instruments." Dilutions of unknown antitoxin are made in 0.85 per cent, sterile salt solution. As a basic dilution one part of the antitoxic serum to nine of the salt solution gives 1/10 c. c. to each cubic centimeter, and from this initial dilution further dilutions may be easily made as follows : 1 c. c. of dilution L + 9 c. c. salt solution = 1-100, etc. A series of mixtures is then made in each of which the quantity of toxin equals the L+ dose, and in which the quantity of antitoxin varies within a wide margin of the limits of strength to be expected. This is illustrated in the following table : L+ (0.29 c. c.) + 1/500 c. c. of antitoxic serum = lives L+ (0.29 c. c.) 4- 1/600 c. c. of antitoxic serum = lives L+ (0.29 c. c.) + 1/700 c. c. of antitoxic serum = lives L+ (0.29 c. c.) + 1/800 c. c. of antitoxic serum = dies in 8 days L+ (0.29 c. c.) + 1/900 c. c. of antitoxic serum = dies in 4 days L+ (0.29 c. c.) 4- 1/1,000 c. c. of antitoxic serum = dies in 2 days 22 Park and Williams. "Pathogenic Bacteria," p. 213. 456 INFECTION AND RESISTANCE In the above tables, according to our previous definition of the antitoxin unit, the serum would contain 900 units to the cubic centi- meter, since 1/900 c. c., injected together with the L+ dose of the standard toxin, resulted in the death of the guinea pig in four days. In order to allow a margin of safety Rosenau and others have sug- gested that the unit should be determined, not by the quantity of antitoxin, which delays death by the L+ dose for four days, but rather by the quantity which, with the L+ dose, results in saving the life of the guinea pig. According to this latter standard the serum employed in the table would be spoken of as containing 700 units to the cubic centimeter. Of course the tabulated measurements M T M MMT BATTERY OF EOSENAU SYRINGES PREPARED FOR ANTITOXIN STANDARDIZATION. (Taken from Kosenau, U. S. Hygienic Bulletin 21, 1905.) are rough, leaving an undetermined zone of 100 units. The exact number of units to the cubic centimeter could, of course, be deter- mined with greater accuracy by now carrying out another series of tests in which the amount of serum varied between 1/700 and 1/900 of a cubic centimeter. In carrying out such a standardization the toxin is diluted so that the L+ dose is contained in 2 c. c. This can easily be done. For instance, in the above the L+ dose being 0.29, it merely necessi- tates adding to each 0.29 c. c. of toxin 1.71 c. c. of salt solution, to each 2.9 c. c., 17.1 c. c. The antitoxin also is made up in such a way that the required dilution is contained in two cubic centimeters, since a total volume of 4 c. c. has been agreed upon as standard for these tests, the injected volume having much influence upon the speed of absorption. In using the so-called Rosenau syringe, shown in the- figure for the standardization of antitoxin, the antitoxin is made up to 1 c. c. in each case, so that 1 c. c. of salt solution may be added to wash out the syringe after injection of the mixture. The mixtures. THERAPEUTIC IMMUNIZATION IN MAN 457 can be made directly in these syringes or in test tubes, and are allowed to stand one hour at room temperature, so that there may be time for complete union. If the mixtures are made di- rectly in the syringes the needles are dipped into sterile vaselin, which closes them and prevents leakage while standing. The mixture is then forced out of the syringe with a rubber bulb, thus ensuring complete injection of all the fluid. As Rosenau states, much depends on the guinea pigs. They must be of standard weight, about 250 grammes, well fed and cared for, and must not be descendants of pigs that have shown marked or unusual resist- ance to diphtheria toxin. This, as Theobald Smith has shown, occasionally happens. The antitoxic serum as obtained from the horse directly may be concentrated in a number of ways, representative of which is the method developed at the Xew York Department of Health by Gib- son,23 Banzhaff, and others.24 The original method consisted in heating horse serum to 56° C. for 12 hours, by which some of the pseudoglobulin was converted into euglobulin, the antitoxin remain- ing in the pseudoglobulin fraction. After this an equal volume of saturated ammonium sulphate solution is added and the globulin precipitated. After several hours the precipitate is filtered off and again taken up in water corresponding in amount to the original volume of serum. After filtration this solution is precipitated with ammonium sulphate and this precipitate is treated with saturated solution of NaCl in quantity twice that of the original serum. After standing for 12 hours the supernatant fluid containing the antitoxin is decanted, and this is precipitated with 0.25 per cent, acetic acid. The resulting precipitate is dried by pressing it between filter papers and is placed in a parchment dialyzing bag, after neutralization with sodium carbonate. At the end of seven or more days of dialyzation against running water, the globulin solution remaining in the dial- yzer is filtered and made isotonic. More recently the method as modified by Banzhaff is as follows : The serum, as obtained from the horse, is diluted by one-half the volume of water, and to this a saturated solution of ammonium sulphate is added up to 30 per cent, saturation. This is heated to 61° C. for two hours. It is then filtered and the residue on the filter paper, which contains the antitoxin, is thoroughly dried by pressing between filter papers and is directly dialyzed. Observations by Park and Throne 25 have shown that this con- centrated antitoxin which, according to Gibson, represents a yield of about 70 per cent, original antitoxic power of the serum, is equally efficient for therapeutic purposes as an unconcentrated preparation 23 Gibson. Journ. of Biol. Chem., Vol. 1, 1906. 24 Gibson and Collins. Journ. of Biol Chem., Vol. 3, 1907. 25 Park and Throne. Amer. Journ. of Medical Science, Vol. 132, 1906. 458 INFECTION AND RESISTANCE and has the advantage of introducing less foreign protein into the human body. It retains its potency, according to Park and Throne, as long as does the whole serum. ACTIVE IMMUNIZATION IN DIPHTHERIA WITH MIXTURE OF TOXIN AND ANTITOXIN Recently Behring 28 has advocated the immunization of human beings with mixtures of diphtheria toxin and antitoxin. This method represents essentially active immunization with toxin ren- dered harmless by neutralization with antitoxin. The use of such mixtures had previously been studied with considerable care, in the case of the toxin of symptomatic anthrax, by Schattenfroh and Grassberger,27 and the procedure had been used in the New York Department of Health for some years in the initial treatment of antitoxin horses. Theoretically considered on the basis of Ehrlich's opinions, one would be inclined to wonder at the fact that relatively neutral mixtures of toxin and antitoxin should possess any antitoxin- inciting properties. Behring explains the immunizing value of such mixtures by the reversible nature of toxin-antitoxin union in the animal body. He calls attention to the fact that our analyses of diphtheria toxin-antitoxin mixtures have been made entirely with guinea pigs as indicators. In studying such mixtures in other ani- mals Behring has come to the conclusion that complete detoxication of the poison in vitro does not occur. He found, for instance, that a toxin-antitoxin mixture that was entirely innocuous for guinea pigs produced an active febrile reaction in an ass. In monkeys (Maca- cus rhesus) he finally found an animal in which he obtained evidence satisfactory to him that toxin may be powerfully active in the animal body, even if it has been previously mixed with antitoxin. If, for instance, he gave a monkey a mixture in which as much as 20 to 40 antitoxin units were mixed with one toxin unit, and repeated the injection two or three times, the animal died of subacute diphtheria toxin poisoning. The mixture ceased to be poisonous for monkeys only when the relation of antitoxin to toxin became one of 80 to 100 antitoxin units to one toxin unit. This final detoxication when suffi- cient amounts of antitoxin were used, it seems to us, may be taken as sufficient evidence that Behring's monkeys did not die of ana- phylaxis. We gather from Behring' s writings that he attributes these dif- ferences in susceptibility to toxin-antitoxin mixtures in various ani- mals to differences in the reversibility of the toxin-antitoxin com- plex in the bodies of the individual species. 26 Behring. Deutsche med. Woch., Vol. 39, No. 19, 1913. 27 Schattenfroh and Grassberger. Deuticke, Wien, 1904; see also Schat- tenfroh, Wien. kl. Woch., No. 39, Sept., 1913. THERAPEUTIC IMMUNIZATION IN MAN 459 Human beings are less susceptible to such mixtures than are monkeys, but nevertheless more so than guinea pigs. It also appears that diphtheria bacillus carriers or such persons who, because of a previous infection, have antitoxin in their blood are much more sus- ceptible to these mixtures than are others. Newborn children are less susceptible than are children from 4 to 15 years. Mixtures which are entirely neutral for the newborn may incite febrile reac- tion in older children. In all cases the injection of such mixtures is followed by a more or less active production of antitoxin. The mixtures which von Behring advocates at present are so prepared that the toxin action upon guinea pigs is practically nil; in other words, the mixture is completely neutralized. The method represents in purpose, and apparently in achieve- ment, a safe process of actively immunizing against diphtheria. Heretofore the method of protecting human beings prophylactically against diphtheria has consisted in the injection of antitoxic serum. This, unquestionably a wise procedure, has nevertheless the disad- vantage of bringing about an immunity of short duration only. Within 20 to 30 days the antitoxin injected may have completely or almost completely disappeared from the blood stream. Prophylactic immunization with the toxin-antitoxin mixtures, however, repre- senting as it does an active immunization, is likely to be more pro- longed in its effects. According to Behring a human being possess- ing 0.01 antitoxin unit in 1 c. c. of blood may be regarded as still moderately protected against diphtheria. According to his estima- tion a decline to this amount, in a person actively immunized by the mixtures (an estimation based upon curve measurements of treated cases), would take about two years. He has observed that horses that had been actively immunized by him, and subsequently used in agricultural work, retained measurable antitoxin values in their blood after five years without treatment. Schreiber28 and others state, also, that this method of active immunization with mixtures of toxin and antitoxin has the advan- tage of avoiding the anaphylactic dangers incident to the injection of antitoxin alone. Their opinion is probably erroneous, since it is most likely that whatever anaphylactic dangers there are result from the injection of horse serum rather than from the antitoxin con- tained in the injected substance. Moreover, the recent studies of Park have shown satisfactorily that the danger of anaphylaxis in the injection of antidiphtheritic sera is practically nil. Among 330,000 cases on record there were but five deaths. The chief value of this new method of immunization is that it represents a safe technique for the prophylactic treatment of indi- viduals exposed to the disease and possibly for the general prophy- lactic immunization of school children, nurses, physicians, etc. In 28 Schreiber. Deutsche med. Woch., Vol. 39, No. 20, 1913. 460 INFECTION AND RESISTANCE the case of children during the ages at which they are most suscep- tible to the disease, the prolonged immunity resulting from the treat- ment should strongly recommend it as a method of promise for the gradual eradication of epidemics. Behring also suggests it as a hopeful method of treatment in the case of bacillus carriers. Schreiber and others have reported upon the effects of treatment when carried out with Behring's mixtures. In the earlier experi- ments of Hahn, mixtures were used in which there was a slight excess of toxin. The later experiments were made with mixtures which were completely neutralized for guinea pigs. In Schreiber's cases from two to six injections were made at intervals of three to five days, most of them subcutaneously, and some of them intramuscu- larly. In no case were there serious reactions, although occasionally there were slight swelling of regional lymph nodes and a little fever. The effects of immunization were noticeable about 23 to 25 days later. When two injections only had been made, at least 0.075 of an antitoxin unit to the cubic centimeter was present. The highest value obtained after two injections was one unit to one cubic centimeter. In nine patients who had been treated by four to seven injections with gradually increasing doses, as much as 10 to 75 antitoxin units to the cubic centimeter resulted. It appears, therefore, that in med- ical practice this method is safe, and that with as little as two injec- tions antitoxin values may be obtained which entirely suffice for the protection of human beings against the ordinary dangers of diph- theria infection, an immunity which, as far as we can judge at present, may last about two years. Another advantage which Behring claims for his method is the production of homologous antitoxin in human beings for the passive immunization of other human beings. Mathes has tried this in children with the idea of thereby avoiding the dangers of anaphy- laxis. Incidentally it was claimed in this case that the passive im- munization, when carried out with homologous serum, lasted longer than did that conferred by horse serum. However, one case is hardly enough to establish such a fact. THE INTRACUTANEOUS METHOD OF DETERMINING TOXIN AND ANTITOXIN VALUES Marks 29 was the first to utilize the prevention of local edema or injury for the determination of antitoxin values. He mixed diph- theria antitoxin and toxin and injected them subcutaneously into guinea pigs, claiming that this method was considerably more deli- cate than the Ehrlich method, since the amount of toxin capable of causing localized edema amounted to as little as one-twentieth of a 29 Marks. Centralbl. f. Bakt., Orig. Vol. 36. THERAPEUTIC IMMUNIZATION IN MAN 461 minimal lethal dose. This method has many points in its favor, and has been recently utilized and improved upon by Homer. Eomer 30 31 32 has developed a method of diphtheria antitoxin standardization which depends upon intracutaneous injections into guinea pigs. The principle of this test consists in the observation that, when very slight amounts of diphtheria toxin are injected intra- cutaneously into the abdominal skin of guinea pigs, small areas of local necrosis result within about 48 hours. When such injections are made with mixtures of toxin and antitoxin the presence of free toxin is indicated by the appearance of such necrosis. Before proceeding to the standardization by this method it ia necessary to determine the "limes-necrosis" (just as Ehrlich deter- mines his L+ dose), that is, the amount of toxin which, together with a given amount of toxin (1/50, 1/200, or 1/2,000), will still produce a minimal amount of necrosis after intracutaneous injection into guinea pigs. It is necessary, therefore, arbitrarily to choose a certain definite fraction of an antitoxin unit and mix this with vary- ing amounts of toxin and inject the mixtures into guinea pigs intra- cutaneously. Those mixtures in which the toxin is fully neutralized will give rise to absolutely no lesion further than, possibly, a slight local edema. Those in which there is a large excess of toxin will cause extensive necrosis. Between the two, in the series, there will be a mixture in which slight local necrosis results from the injection. In this mixture the amount of toxin, just sufficient to cause notice- able necrosis in spite of admixture with the antitoxin, contains the L-n (limes necrosis) dose. When this has been determined, then unknown antitoxin can be similarly measured against this L-n dose of the standard toxin. The method has the advantage of permitting one to work with very small quantities, since only a small fraction of a cubic centimeter need be used for intracutaneous injections ; also it permits great economy of animal material, since four or five tests can be simultaneously car- ried out upon the abdominal wall of the same guinea pig. The technique is not easy. We have found in studying this method in connection with some work carried on in our laboratory by Dr. M. C. Terry, that a considerable amount of practice and experi- ence is necessary, both in carrying out the procedure accurately and in judging the lesions. However, when carefully and consistently done by an experienced worker, this method gives results which cor- respond with fair accuracy to measurements made of the same anti- toxin by the Ehrlich method. This has been the experience of Lewin,33 and also of Terry in the few experiments carried out by him, 30 Romer. Zeitschr. f. 1mm., Vol. 3, p. 208, 1909. 31 Romer and Sames. Ibid., p. 344. 32 Romer and Somogyi. Ibid., p. 433. 33 Lewin. Centralbl. f. Bakt., Orig. Vol. 67, 1913. 462 INFECTION AND RESISTANCE The Homer method has been recently used by clinicians for the" determination of the presence of free toxin or antitoxin in the circu- lating blood of patients suffering or convalescent from diphtheria. Romer himself suggested this, since his method is adapted to the determination of extremely slight amounts of either substance. A recent study by Harriehausen and Wirth 34 illustrates the results obtained in such tests. Normal human serum injected intracuta- neously into guinea pigs never caused necrosis. Neither did the similar injection of the sera of children suffering from varicella and other diseases. Of twelve children suffering from diphtheria, how- ever, serum taken before the administration of antitoxin caused necrosis upon intracutaneous injection into guinea, pigs, in every case. In spite of the administration of antitoxin, toxin was demon- strable in the blood in five cases. as long as the 35th day. Of ten cases of post-diphtheritic paralysis, toxin was demonstrated in the blood of five. Since this method of determining antitoxin values in the blood of human beings is of considerable importance and may have much practical value, it may be useful to insert an example of such an application of this method as used by Hahn 35 in a series of investi- gations mentioned elsewhere. The standard toxin was obtained from Marburg. In a series of guinea pigs a determination was made of the smallest quantity of this standard poison which would produce just noticeable necrosis of the skin if injected into the pig intracutaneously, together with 1/2, 000th of a unit of a standard antitoxin. The toxin and antitoxin were left together for 24 hours before injection, 3 hours in the incu- bator, and 21 hours in the refrigerator. When this quantity of the antitoxin had been determined, it could be used in similar experiments and similarly mixed with vary- ing amounts of the patient's serum. The amount of antitoxin pres- ent in such serum could then be easily computed. For, let us sup- pose that this amount of toxin, together with 1/5 00th of a c. c. of the serum injected intracutaneously into the guinea pig, gave the same amount of necrosis in the same time as the identical quantity of the toxin, together with 1/2, 000th of a standard unit. Then l/500th of a patient's serum was equivalent to 1/2, 000th of a stand- ard unit, and the patient's serum would contain 0.25 of a unit per cubic centimeter. Michiels and Schick 36 have carried out intracutaneous reactions with diphtheria toxin directly upon the human body to determine whether or not diphtheria immunity was present. They injected 0.1 c. c. of a 1 to 1,000 dilution of toxin and claim that a positive 34 Harriehausen and Wirth. Zeitschr. f. Kinderheilkunde, Vol. 7, 1913. 35 Hahn. Deutsche med. Woch., Vol. 38, No. 29, 1912. 36 Michiels and Schick. Zeitschr. /. Kinderheilkunde, Vol. 5, 1912. THERAPEUTIC IMMUNIZATION IN MAN 463 intracutaneous reaction with this amount indicates an absence of antitoxin from the blood, or at least an insufficient protection. The Schick reaction is at present carried out at the New York Depart^ ment of Health, under the direction of Park, with 1/5 Oth M L D intracutaneously injected. The dilutions are so made that this quantity is contained in a total volume of 0.1 c. c. TETANUS ANTITOXIN AND ITS STANDARDIZATION The methods employed in the production and standardization of tetanus toxin are in every way analogous to those used in the case of diphtheria antitoxin. A strong toxin is obtained by growing the organisms under anaerobic conditions on suitable media. Accord- ing to Yaillard and Vincent37 it is essential that the media upon which the tetanus bacilli are grown should be freshly made and sterilized. Apparently this precaution, which has been similarly recommended by Wladimiroff, Novy, and others, is made necessary- by the gradual absorption of oxygen which takes place if the media are allowed to stand for a long time without heating. It is further necessary in preparing tetanus toxin that the culture medium should not be acid, and a weakly alkaline initial titre is advised. For the same reason, also, most workers have advised against the use of glucose or other carbohydrates in the media, since the acid formed by the fermentation of these substances inhibits growth and toxin production. Eecently Hall 38 has advised the use of a simple meat extract broth to which have been added 1 per cent, of dextrose and 0.5 per cent, of finely powdered magnesium carbonate. The last- named substance, by neutralizing any acid that is formed from the glucose, prevents the harmful acidity. Anaerobic conditions are ob- tained by growing the organisms under a layer of oil in tightly stop- pered flasks. Although mice were formerly used in the standardization of tetanus toxin and antitoxin, the more recent usage has been to sub- stitute guinea pigs as in diphtheria standardization. According to the recent directions of Rosenau and Anderson 39 the purposes of the standardization are carried out as follows: The unit of antitoxin is arbitrarily designated as 10 times the smallest amount of serum necessary to preserve the life of a guinea pig weighing 350 grams for 96 hours, when given together with an official test dose of toxin. The test dose of toxin contains 100 min- imal lethal doses. And the minimal lethal dose is measured against a 350-gram guinea pig. 37 Vaillard and Vincent. Ann. Past., 1891. 38 Hall. "Univ. of Cal., Publ. in Path.," Vol. 2, No. 11, 1913. 39 Rosenau and Anderson. U. S. P. H. Service Hyg. Lab. Bull. 43, 1908. 464 INFECTION AND RESISTANCE In carrying out the standardization the L+ dose of 'the toxin is used, but, unlike diphtheria standardization, in this case the L+ dose means an amount of toxin which will kill a guinea pig of 350 grams in four days, although united with 0.1 unit of antitoxin (it must be noted that the L+ dose in this case is measured against one- tenth unit of antitoxin rather than against 1 unit, as in the case of diphtheria. In determining the value of an unknown antitoxin, mixtures are made, each containing the L+ dose of the toxin and varying quan- tities of antitoxin. As in diphtheria measurements, the various in- jection volumes are brought to 4 c. c. with salt solution, and are then injected subcutaneously into guinea pigs of about 350 grams. The table given below is taken from the Bulletin of Rosenau and Ander- son. Subcutaneous injection of a mixture of No. of Weight of guinea guinea pig Time of death Pig (grams) Toxin (Test dose) Antitoxin ((, * \ (gram) 1 360 0.0006 0.001 2 days 4 hours 2 350 0.0006 0.0015 4 days 1 hour 3 350 0.0006 0.002 Symptoms 4 360 0.0006 0.0025 Slight symptoms 5 350 0.0006 0.003 No symptoms In this experiment 0.0015 equals 0.10 antitoxin unit. ANTITOXINS AGAINST SNAKE POISONS (Antivenin) Antitoxins against snake poisons have been produced by a num- ber of different workers, but the subject has been most extensively studied by Calmette. As early as 1887 Sewall 40 succeeded in in- creasing the resistance of pigeons to snake poison. Later Calmette and Physalix and Bertrand independently succeeded in producing immunity in rabbits and guinea pigs with the poison of the cobra. The serum of animals treated with snake poisons gradually acquires antitoxin properties, but the process of immunization is not a simple one, and considerable time is needed for the immunizations. Snake poisons, as we have seen, have attracted considerable atten- tion because of their peculiarities in being antigenic and yet differ- ing in heat resistance and a number of other properties from the 40 Sewall. Cited from Calmette. THERAPEUTIC IMMUNIZATION IN MAN 465 bacterial toxins. It was with snake poisons that Calmette definitely showed that the union of toxin and antitoxin is a true neutralization and is not accompanied by the destruction of the toxin. These ex- periments, as we have seen, were elaborated later by Morgenroth, who succeeded in producing the snake poison HC1 combination. It is these poisons also that have been the subject of extensive study by Flexner and Xoguchi, by Kyes, and later by von Dungern and Coca. This work has been sufficiently discussed in other places and need not occupy us here. The important poisonous snakes may be divided into the colubridse, to which class the cobra belongs, and the viper- idse, which includes the ordinary European vipers, the rattlesnake, and most of the poisonous snakes of North and South America. Ac- cording to Calmette the poison of the cobra is much more heat-stable than that of the rattlesnake. Pharmacologically the poisons of these two main classes of snakes show considerable difference. In the case of the cobra there is very little local disturbance and the systemic symptoms dominate the clinical picture. Calmette describes the cobra bite as being followed only by a feeling of stiffness at the site of the bite, followed very soon by great general weakness, difficulty in respiration, slow heart action, and finally death with unconscious- ness. In the case of the vipers the local symptoms are very much more marked, there being great pain and swelling and apparent clot- ting of the blood about the point of the bite, with a rather slower onset of systemic symptoms. In a description by Sparr 41 of a case of bite by Russell's viper there was almost immediate swelling of the limb with a faint bluish tint around the pin-point puncture, and within 15 minutes great weakness, restlessness, and retching. In spite of very active local treatment, within a short time after the bite, the patient died within 24 hours of asphyxia and heart failure. According to Calmette 0.0002 gm. of cobra poison will kill a guinea pig; Noguchi states that 0.0005 gm. of rattlesnake venom will kill a guinea pig of 250 gr. within 24-30 hours when injected intra- peritoneally. The snake poisons apparently contain substances which are especially active upon nerve cells (neurotoxins), and hemolysins which act particularly upon the red blood cells. Flexner and Noguchi 42 also speak of another poison which acts particularly upon the endothelium of the blood vessels producing hemorrhages. According to Calmette the antisera which are produced by im- munization with cobra poison are most strongly potent against neuro- toxic poisons of the colubrida3 and, to a certain extent, against some of the poisons of the vipers. However, the action of the cobra anti- toxin against viper poison seems at best to be weak. On the other hand, antitoxins produced with rattlesnake poison are not potent against the cobra venom since, as Calmette states, the rattlesnake 41 Sparr. Biochem. Bull, Dec., 1911, No. 2. 42 Flexner and Noguchi. Univ. of Pa. Bull, Vol. 15, 1902. 466 INFECTION AND RESISTANCE poison contains hardly any neurotoxin. Antitoxins may be produced by the gradual immunization of horses, and have been produced in this way by Calmette in the Pasteur Institute of Lille for some years. Calmette standardizes his antitoxin by determining the amount of serum which completely neutralizes in vitro 0.0001 gm. of the poison as tested upon white light. He also determines the protective power by injecting a rabbit with 2 c. c. of the serum and two hours later gives 1 gm. of the poison. Noguchi has studied rattlesnake poison particularly and suc- ceeded in preparing a strong antitoxin by the gradual immunization of a goat. Great difficulty has always been experienced in attempts at immunization with rattlesnake poison because of the very violent local injury produced by injections of the venom. The potency of the serum produced by him was such that 2% c. c. of goat serum protected guinea pigs against 12 times the fatal dose of rattlesnake poison if given at the same time. If the antivenin was given one hour later, 5 times the amount of serum had to be given. PASSIVE IMMUNIZATION IN DISEASES CAUSED BY BAC- TERIA WHICH DO NOT FORM SOLUBLE TOXINS As we have stated in another place the greatest therapeutic suc- cesses with passive immunization have been achieved in bacterial diseases in which the malady is essentially a toxemia due to a soluble toxin. In such cases the serum of actively immunized animals con- tains specific antitoxins by virtue of which the toxins circulating in the blood of the patient are directly neutralized, quantity for quan- tity, with consequent therapeutic benefit. In the case of bacteria in which no toxins are formed, the immunization of an animal is not followed by the formation of any poison-neutralizing principle. Here the injection of bacteria, dead or alive, or the invasion of the bacteria in the course of spontaneous disease, is followed by the formation of specific antibacterial substances, lytic, opsonic, agglu- tinating, or precipitating bodies, the nature of which we have dis- cussed in other chapters. The toxemia which occurs in such cases is due as we have seen to derivatives of the bacterial protein which by some observers are regarded as preformed endocellular poisons liberated by the lytic action of the serum, and by others as split products of the bacterial protein, non-existent until the bacterial cell has been acted upon by the serum components and destroyed. How- ever this may be, the recovery from diseases of this nature is accom- plished by bacterial destruction ; this may be direct, by the bacteri- cidal action of the serum, or indirect by opsonic properties which induce phagocytosis. The poisons which are liberated from the bac- terial bodies, if free, can do their injury, and no neutralizing sub- THERAPEUTIC IMMUNIZATION IN MAN 467 stance is formed in the body fluids to prevent their action as far as we know. Immunity in such cases, then, is not an antitoxic immu- nity in any sense of the word ; it is rather an antibacterial immunity in which the disease is prevented or cured only when complete de- struction of the bacteria has taken place. If an animal or a human being is prophylactically immunized against diseases of this kind (typhoid, cholera, etc.), it is easy to see that an increased presence of antibacterial substances, bactericidal or opsonic, in the circulation would serve efficiently and rapidly in disposing of the small numbers of invading micro-organisms which ordinarily enter the body in spon- taneous infections. And, indeed, experience has shown that prophy- lactic immunization can be successfully carried out in the case of cholera, typhoid fever, plague, and other diseases which are suffi- ciently prevalent endemically or epidemically to justify prophylaxis on an extensive scale. However, when in diseases of this kind the body is already exten- sively infected and has begun, as is usually the case, to respond spon- taneously with the formation of specific antibodies, it has been a matter of doubt whether or not passive immunization, that is, the introduction of specific antibodies in the form of the serum of a highly immunized animal, is therapeutically of value. Indeed, it has been feared that the use of such sera may even be harmful in that the sudden introduction of large amounts of bactericidal sub- stances might lead to a sudden liberation of large quantities of poi- sonous products and consequent rapid toxemia. The conditions in such cases are exceedingly complex and many gaps exist in our knowledge concerning them. The bacteria when invading the body, immediately enter into conflict with the protective forces, as we have stated in the chapter on Infection. If a consider- able degree of resistance exists, let us say as the result of preceding immunization or a recent attack of the disease, there is a rapid destruction of the bacteria, probably by active phagocytosis. It has been shown by Bordet in the case of cholera and more recently by Gay with typhoid, that injection of the organisms into immunized animals is followed by prompt and high leukocytosis, whereas sim- ilar injections into normal animals usually induce a temporary leuko- penia. When the invaded animal is not particularly resistant the bacteria may accumulate and, as in the case of pneuraococci and streptococci, develop phagocytosis-resisting properties (capsule forma- tion, etc.) ; or, as in the case of typhoid bacilli, there may be an immediate liberation of toxic substances (anaphylatoxins) by reac- tion between bacterial cell and blood plasma, wrhich can induce leuko- penia, and by this means the organisms may be protected from phagocytic destruction. Experience with curative sera in all of the conditions of this class has yielded promising results only when the cases have been treated with the sera at early stages of the disease, 468 INFECTION AND RESISTANCE either when the invading germ was still localized or, at least, wnen the septicemic condition was not yet thoroughly established. It may be that the doses heretofore given have been insufficient, and indeed recent experiences with pneumonia seem to indicate that this may have been, in part, the cause of earlier failures. Yet in pneumonia the septicemia probably does not represent the firm establishment of a foothold by the pneumococcus in the circulation but rather a con- tinuous discharge of new organisms into the blood from the localized lesion in the lung. It is our own opinion moreover that septicemia as usually ob- served clinically represents in most cases exactly this condition, that is, a more or less continuous discharge of the bacteria into the blood from some active focus with a continuous destruction of the organ- isms after they have entered the blood stream. It is only when the resistance of the body is overwhelmed, in the later stages of the disease, that the bacteria can continue to grow and develop in the circulation, and this stage probably does not occur until death is imminent. In such septicemic diseases as streptococcus infection, typhoid fever, plague, anthrax, and many others the presence of the bacteria in the blood at the time when the patient is still in a condi- tion of powerful resistance probably means that the bacteria are being supplied to the blood from the local lesions. There is prob- ably just such a continuous discharge of bacteria from the focus into the blood with active destruction after the bacteria have entered the circulation. This seems especially probable from the fact that in many of these diseases the protective antibodies, bactericidal and opsonic, can often be demonstrated in the blood serum in quantities higher than normal at the very time when blood culture yields posi- tive results. In typhoid fever, of course, it is well known that bac- tericidal titres of over 1-50,000 are often present while the patient may still be very sick, and in the more chronic streptococcus condi- tions with malignant endocarditis we have often seen that opsonic properties on the part of the patient's serum against the very organ- ism invading him are considerably higher than normal. We take this to mean that the injection of immune sera would simply aid in more rapidly freeing the blood stream of the bacteria, the cure of the disease, however, involving a destruction of the focus. This, of course, is not possible merely by the injection of the serum. When, as in some cases of streptococcus infection, the focus can be surgically reached, the septicemia will often disappear and cure result, as we have ourselves had the opportunity to observe. When the focus cannot be reached surgically, it may nevertheless be a wise procedure to inject considerable amounts of immune serum, for, by keeping the blood stream free of bacteria, the case may be influenced favorably. Pneumonia is an example of this. Former failures have recently been turned into partial success by the work of Neufeld and of Cole THERAPEUTIC IMMUNIZATION IN MAN 469 merely by the use of larger quantities of immune sera essentially similar to sera used at previous times, and Cole attributes the appar- ently favorable results to the fact that the blood stream can be cleared of bacteria although the focus cannot itself be affected. Cure of such diseases, therefore, by serum treatment can hardly be expected. Favorable influence of the disease by energetic serum treatment may, however, be hoped for. In discussing this subject it must not be forgotten, however, that in most of the diseases which we have classified, on the basis of pre- vailing opinions, as caused by bacteria that do not form true toxins, the formation of such poisons has been claimed by a number of care- ful and eminent observers. In the case of the typhoid bacillus, espe- cially, Chantemesse, Kraus and Stenitzer, and others have claimed the existence of a true toxin and a consequent antitoxin in immune sera. Similar claims have been made for the cholera spirillum by Kraus and Doerr, for the streptococcus by Marmorek, and for the plague bacillus by Markl and Rowland. Since these claims have been made on the basis of extensive experimentation by competent men the question must be left open, and the possibility of antitoxic properties on the part of the sera cannot be completely ignored. Since in most cases, however, the poison-neutralizing properties of the immune sera in this disease have not exceeded more than 1 to 2 multiples of the M L D of the bacterial poisons, it does not seem impossible that the apparent antitoxic properties may have repre- sented merely an acquired tolerance to anaphylatoxic poisons of which we have spoken in another place. SERUM TREATMENT IN EPIDEMIC CEREBROSPINAL MENINGITIS Serious attempts to produce curative sera against the epidemic form of cerebrospinal meningitis were not made until 1906 and 1907, when this disease appeared epidemically chiefly in Europe, where it appeared most severely in Eastern Germany, and in the Eastern United States. In 1906 Kolle and Wassermann immunized three horses with meningococci, using for immunization purposes the dead organisms followed by living cultures and cultures shaken up in distilled water, the so-called artificial aggressins of Wassermann and Citron. They obtained sera of considerable potency when measured against menin- gococcus cultures, and suggested standardizing the sera by comple- ment fixation. They did not at this time treat human beings, but sug- gested the use of the serum subcutaneously and intravenously in meningitis cases. Very soon after the publication of the work of Kolle and Wassermann Jochmann 43 also produced an antimeningo- 43 Jochmann. Deutsche med. Woch., 1906, Vol. 32, p. 788. 470 INFECTION AND RESISTANCE coccus serum by immunizing horses with proved meningococcus cul- tures, in his cases making a polyvalent serum by the use of many different strains of the organism. The sera which he obtained were highly agglutinating, somewhat bactericidal, and, according to him, not antitoxic. He first succeeded in immunizing guinea pigs against meningococci by injecting the serum 20 hours before infecting the animals. He also treated 40 cases of meningitis in man and ob- tained encouraging results in cases treated before the development of hydrocephalus. Believing that possibly intraspinous injection of the serum might offer advantages, he first determined by experiments upon the dead body that the injection of methylene-blue intra- spinously passed from the point of injection in the lumbar regions as far up as the olfactory nerves. After having determined this he treated 17 cases by tapping the spinal canal, taking out 30 to 50 c. c. of spinal fluid and then injecting about 20 c. c. of the serum. Of these 17 cases only 5 died, and Jochmann expresses himself opti- mistically in consequence. Meanwhile Flexner 44 had been working upon the same subject, laying a rather more thorough basis for therapy in careful animal experimentation. He produced the typical disease in monkeys by intraspinous inoculation of the meningococci and then saved the animals from death by following the infection with the injection of serum intraspinously six hours later. In his earlier articles he ex- presses himself with much conservatism, but his studies were con- tinued and extensive opportunity for testing the serum which he then produced, together with Jobling,45 was offered by the continuance of the epidemic throughout the United States. The results with the serum produced at the Rockefeller Institute have since then proved to be uniformly favorable. The method of intraspinous inoculation of the serum after the removal of some of the spinal fluid was the method finally adopted by Flexner as most favorable, and this is the method in current use to-day. In 1908 Flexner and Jobling reported upon 47 cases treated with the anti- serum of which 34 recovered. Of 12 additional cases reported in an addendum only 4 died. In the most recent summary by Flex- ner 46 records of 1,294 cases that have been treated with the serum prepared at the Rockefeller Institute are analyzed. Of this num- ber, unselected and treated in many different parts of the world, 69.1 per cent, recovered. It is of course very difficult to obtain exact comparative data on the efficiency of any method of treatment in a disease as irregular in its clinical manifestations as epidemic meningitis, especially since the mortality attending upon different 44 Flexner. J. Exp. Med., Vol. 9, 1907, and /. A. M. A., 1906, Vol. 47, p. 560. 45 Flexner and Jobling. J. Exp. Med., Vol. 10, 1908. 46 Flexner. J. Exp. Med., Vol. 17, 1913. THERAPEUTIC IMMUNIZATION IN MAN 471 epidemics is subject to great variations. For this reason we can draw conclusions only from a large statistical material. However, we know that the average mortality of epidemic meningitis before the introduction of specific therapy ranged certainly higher than 65 per cent., and in carefully studied epidemics usually between 70 and 80 per cent. The statistics of Flexner showing a mortality hardly exceeding 30 per cent, in unselected cases unquestionably marks a wonderful therapeutic triumph. It must be remembered in consider- ing the benefits of this serum that in unselected cases there must be many in whom the disease has produced marked anatomical changes in the central nervous system before the serum is used. It is well known, of course, that the later manifestations of this disease, which often lead to death with hydrocephalus, asthenia, and malnutrition, are the remote results of the anatomical injuries produced by the in- flammatory reactions accompanying the earlier manifestations of the acute infection. These conditions of course cannot be expected to yield to serum treatment. It must be assumed, therefore, that were we able to obtain statistics of cases diagnosed and treated soon after the onset the figures would be even more favorable than those stated above. The action of the serum seems very largely to be an opsonic one, in that, under the influences of serum, a powerful phagocytosis of the meningococci takes place. It is also possible that to a certain extent bactericidal action participates, in that the injection of the serum into the closed space may give rise to a sort of intraspinous Pfeiffer reaction with energetic ingestion of the bacteria by leu- kocytes. The standardization of the antimeningococcus serum has been worked out particularly by Jobling.47 After attempting to stand- ardize the sera by their protective power against meningococcus in- fection in animals and by complement fixation, as suggested by Kolle and Wassermann, Jobling has come to the conclusion that neither of these methods is sufficiently regular, and that the most suitable procedure is a standardization by opsonin determination. The method as worked out by him depends upon determining the highest dilution of the immune serum at which opsonic action against the meningococcus is still evident. He suggests as a definite and suitable standard of strength opsonic activity at a dilution of 1-5,000 of the antiserum. SERUM TREATMENT IN STREPTOCOCCUS INFECTIONS The attempts to produce powerful immune sera against strepto- cocci date back to the earliest days of immunology. That the sub- ject is a particularly difficult one follows from the great confusion 47 Jobling. J. Exp. Med., Vol. 11, 1909. 472 INFECTION AND RESISTANCE which has prevailed, and, to a great extent, still prevails, regarding the classification of the streptococci and their interrelationship. There are apparently a large number of different strains of strepto- cocci which vary from each other, not only culturally, but also in regard to agglutination and bactericidal reactions. For this reason it is not at all a foregone conclusion that a serum prepared by the immunization of an animal with a streptococcus of one type will have any protective action against other strains. The subject has been still more complicated recently by the discovery of Rosenow 48 that the various types of streptococci (viridans, hemolyticus, etc.) are not constant in their properties, but may be artificially trans- formed one into the other, and that even mutation of true pneumo- cocci into true streptococci may take place. Most important in this connection is the observation that a pneumococcus sent to Rosenow was altered by him by special methods of cultivation in such a way that not only its morphological and cultural properties were changed, but also its agglutination reactions. These observations are of the utmost importance in connection with attempts at producing specific sera which can be utilized therapeutically. In all cases, therefore, in which streptococcus immune serum is at all used it must be re- membered that the disease produced in human beings by organisms classified among the streptococci are by no means necessarily closely related in biological reactions, and the same immune serum may be extremely potent in one case and entirely useless in another. That animals could be successfully immunized against strepto- cocci was shown early in the history of investigations in immunity by a number of workers, notably Roger, Behring, von Lingelsheim, and Mironoff. The first extensive attempts to produce a curative serum for use in passively immunizing human beings were made by Marmorek 49 at the Pasteur Institute in 1895. The basic idea from which Marmorek worked was the similarity of all the streptococci producing disease in human beings. He also believed that the most powerful serum could be produced with cultures whose virulence had been greatly enhanced by animal passages. When such cultures were grown on mixtures of human serum and broth he asserted furthermore that soluble poisons were produced which could be ob- tained by filtration of the culture fluids. For these reasons he im- munized horses with cultures rendered highly virulent by very gradual injections first of dead then of living organisms, finally injecting also considerable quantities of culture filtrates. Testing these sera upon animals, he was successful in protecting against streptococcus infection when the serum was administered 12 to 18 hours before the bacteria were injected. He expressed the opinion that the serum was antitoxic as well as antibacterial. In 48Rosenow. Journ. A. M. A., Feb., 1914. 49 Marmorek. Ann. Past., Vol. 9, 1895. THERAPEUTIC IMMUNIZATION IN MAN 473 his earliest reports the results of the treatment of 413 cases of ery- sipelas leave one very much in doubt as to the value of the serum since the difference in mortality between the treated and the untreated cases is less than 2 per cent. However, an analysis of the individual cases makes the serum treatment appear more favorable. He re- ported good results also in 7 cases of puerperal septicemia and in scarlatinal angina. Later observers, notably Lenhartz,50 Baginsky,51 and many others, have not been able to confirm the favorable results reported by Marmorek, and it may be stated that at the present day the value of Marmorek's serum is very much in question. Anti- streptococcus sera have also been produced by Aronson 52 and Tavel, Van de Velde, Menzer,53 Moser, 54 and some others. Aronson at first worked from the idea which Marmorek also had used that there was a close relationship between the various streptococci pathogenic for man. He adopted the opinion first developed by Denys 55 and Van de Velde that many different strains should be used for im- munization in order to allow for possible difference in the character- istics of the pathogenic streptococci. This principle of the necessity for the production of polyvalent sera was also emphasized strongly by Tavel, who based his opinion on careful agglutination tests, and by Menzer and Moser. That the action of the antistreptococcus sera, however produced, is very largely due to its opsonic properties has been shown by Bordet,56 by Meier and Michaelis, and a number of other workers. If there is any bactericidal power it is probably relatively slight. It would be quite impossible to attempt in this place to analyze the large number of streptococcus infections of man which have been treated with one or the other antistreptococcus sera. Those men- tioned, moreover, do not by any means include all the sera which have been produced and marketed for this purpose. In general we may say that here, as well as in the cases of other sera in which no antitoxic action is evident, beneficial results have been obtained chiefly in cases in which the streptococcus infection has been localized and treated early after its inception. In generalized or advanced cases it cannot be said that the results are encouraging. Even in animals, in which experimental conditions can be so much more thoroughly controlled, the protective action of even the strongest sera is evident only if the serum is administered either before in- 50 Lenhartz. "Die Septischen Erkrankungren Holder," Wien, 1903. 51 Baginsky. Berl kl Woch., 1896, p. 340. 52 Aronson. Berl. kl. Woch., Vol. 39, 1902; Deutsche med. Woch., 29, 1903. 53 Menzer. Berl. kl Woch., 1902, and Munch, med. Woch., 1903. 54 Moser. Wien. kl. Woch., 1902. 55 Denys. Bull, de VAcad. Beige, 1896, cited from Schwoner K. and L. H., Vol. 2. 56 Bordet. Ann. de I'Inst. Past., 1897. 474 INFECTION AND RESISTANCE fection or within a very definite period after inoculation. The standardization of streptococcus sera may be accomplished by de- termining its protective value for animals when injected 18 to 20 hours before infection. When the sera are produced by immuniza- tion with streptococci obtained from the human body and without pathogenicity for animals the standardization is of course unsatis- factory. SEBUM TREATMENT IN PNEUMONIA Attempts to work out a therapeutically valid method of passive immunization in pneumonia have been many and date from the very beginning of the discovery that pneumonia was a bacterial infection. Sera have even been marketed and used, but until recently no very encouraging results were obtained. Kecent studies have revealed that in pneumonia the serum of convalescents contains practically no bactericidal properties for the pneumococcus, and that the protective powers of such serum depend upon the presence of immune opsonins or bacteriotropins, by means of which the pneumococci are ren- dered amenable to phagocytosis. Virulent pneumococci are not as a rule phagocytable in the presence of normal serum. However, in the presence of immune serum powerful phagocytic action can be observed. Neufeld has studied the conditions of pneumococcus immunity most thoroughly in recent years. The most important advance from a practical point of view was a discovery made by him, with Han- del,57 in 1909. They determined that there was a definite difference between various pneumococci in their reactions to immune serum ; in other words, pneumococci could be grouped into various serological types. The serum produced with organisms of one type did not pro- tect against infection with other strains. In consequence they called attention to the importance of determining the type of pneumococcus which causes the individual pneumonia so that the corresponding immune serum might be used. They produced a highly potent anti- pneumococcus serum by the injection of horses and donkeys with highly virulent pneumococci grown on fluid cultures, then deter- mined the high protective power of this serum upon animals and used it in the treatment of patients by intravenous injection. Their results were exceedingly encouraging. In reporting their results Neufeld and Handel state that considerable doses must be given. They call attention to the fact, revealed by their animal experiments, that moderate amounts do not, as in the case of diphtheria serum, exert a correspondingly slight amount of beneficial action, but that in the case of the pneumonia serum amounts smaller than a certain 57Neufeld and Handel. Zeitschr. f. Imm., Vol. 3, 1909, and Arb. aus dem kais. Gesundh. Amt., Vol. 34, 1910. THERAPEUTIC IMMUNIZATION IN MAN 475 active minimum seem to exert absolutely no beneficial action. This is a fact which later was also determined by Dochez. In confirmation of the work of Neufeld and Handel 58 Dochez and Gillespie 59 have also been able to determine that there are at least two distinctive groups of pneumococci which differ from each other as far as agglutination and serum protection experiments are concerned. In addition to these two fixed types they separate as a third group the streptococcus or Pneumococcus mucosus and a fourth heterogeneous group which seems to fit in with none of the others as far as serum reactions can determine. Cole,60 therefore, adopts the reasoning of Neufeld in that he advises the determination of the type of pneumococcus present in cases of pneumonia as a guide to the type of antiserum to be used. The type of organism is determined as soon as a case comes under observation and 50 to 100 c. c. of the homologous antiserum is in- jected intravenously. The result in the few cases so far treated by Cole and his associates has been encouraging. Protective substances, according to Dochez, appear in the serum of treated cases very shortly after the administration of the serum, whereas in untreated lobar pneumonia such protective substances usually do not appear until after the crisis. Apparently Cole believes that the great value of passive immunization of this kind in pneu- monia lies in the fact that the bacteriemia shown to prevail in prob- ably all cases of lobar pneumonia is either cured or improved by the treatment, converting the disease, which is by nature, at least for a time, a septicemia, into a localized pulmonary infection. Experi- ence with antipneumococcus serum so far has been too limited to warrant final judgment as to its permanent place among therapeutic agencies. THE SERUM TREATMENT OF TYPHOID FEVER The first extensive attempts to treat typhoid fever by passive im- munization with the serum of treated animals were made by Chante- messe, who immunized horses with filtrates of typhoid cultures sub- cutaneously, and with emulsions of virulent bacilli intravenously. Chantemesse believed that the serum of horses which had been treated in this way for very long periods possessed, not only bacteri- cidal action, but stimulated phagocytosis, and possessed a certain limited amount of neutralizing power against the toxic properties of the typhoid filtrates. At the International Congress of Hygiene in Berlin in 1907 Chantemesse61 reported upon a thousand cases 58 Neufeld and Handel. Arb. aus dem Jcais. Gesundh. Amt., Vol. 34, 1910. 59 Dochez and Gillespie. Jour. A. M. A., Vol. 61, p. 727, 1913. 60 Cole. Jour. A. M. A., Vol. 61, p. 663, 1913. 61 Chamtemesse. International Congress of Hygiene, Berl., Sept., 1907 ; Ref. Bull, de rinst. Pasteur, Vol. 5, 1907, p. 931. 476 INFECTION AND RESISTANCE treated with his serum. Of this number 43 only died, whereas the average mortality during the same six years at the Paris hospitals was IT per cent. The injection of the serum he claimed very mark- edly improved the condition of patients in that, after a preliminary period of no apparent change lasting from several hours to 5 or 6 days, the temperature goes down and the general condition of the patient changes considerably for the better. He noticed very few complications in these cases, and intestinal hemorrhage occurred four times only. A remarkable feature of Chantemesse's treatment is that he in- jected into the patients a few drops only of the serum, and rarely made a second injection, facts which alone tend to persuade one that his apparent therapeutic success was a fortunate accident. The opinion originally expressed by Chantemesse that the serum of horses vigorously treated with typhoid bacilli possesses in addition to its bactericidal and opsonic powers definite antitoxic properties recurs again in the work of a number of investigators. Besredka 62 prepared a serum by the intravenous injection of typhoid cultures heated to 60° C., continuing the immunization for 6 months. He claims that this serum possesses what he designates as "anti-endo- toxic" properties. A dry extract of typhoid bacilli which in dose of 0.01 gram killed a guinea pig of 300 grams regularly became innocu- ous when mixed with small quantities of this horse serum. One c. c. of the horse serum neutralized often as much as two fatal doses of the serum, but it is important theoretically to recognize that Besredka states particularly that even an increase of the quantity of serum never neutralized more than two fatal doses. This is particularly important in connection with the more recent studies on toxic split proteins by Vaughan, and on anaphylatoxins by Bessau and by Zins- ser and Dwyer, in which it has been shown that an animal acquires a tolerance against the toxic substances produced from bacterial and other proteins which, however, never exceeds one or two multiples of the minimum lethal dose. This fact alone would militate against considering the serum of Besredka in any way antitoxic in the sense in which the word is used concerning diphtheria and tetanus anti- toxins where neutralization of poison follows roughly the law of multiples. Besredka's anti-endotoxic sera has recently been very thoroughly investigated by Pfeiffer and Bessau.63 These investi- gators have found that Besredka's serum exerted a very definite beneficial influence upon typhoid infection in guinea pigs if injected at the same time with the bacilli. In their experiments it also pro- tected somewhat against the toxic properties of substances derived from the typhoid bacillus, and Pfeiffer and Bessau did not believe that this was due to a true antitoxic action, nor that the serum was 62 Besredka. Ann. de I'Inst. Pasteur, 19, 1905, and 20, 1906. 63 Pfeiffer and Bessau. Centralbl. f. Bakt., Vol. 56, 1910. THERAPEUTIC IMMUNIZATION IN MAN 477 superior in this respect to the ordinary bactericidal sera prepared by inoculating animals with typhoid bacilli. Kraus and Stenitzer64 have also taken up the study of typhoid immunization from the point of view that the typhoid bacillus produces a true toxin, and that therefore a true antitoxic action could be expected from the sera pro- duced by immunization with typhoid filtrates. It should be noted that, in spite of the most common opinions against this at present, a similar point of view was advanced by MacFadyen,65 and more recently by Arima.66 Kraus and Stenitzer 67 immunized horses and goats very highly with extracts of agar cultures and with broth fil- trates by intravenous injection. The serum which they produced in this way not only possessed the ordinary bactericidal action, but, they claimed, neutralized also toxic broth filtrates, not only of the typhoid, but of the paratyphoid bacilli. The serum of Kraus and Stenitzer has been used by a number of observers, among whom are Herz,68 Forssmann, linger, Russ, and others, and the results are said to be encouraging in early cases. Rodet and Lagrifoul 69 immunized horses with living typhoid cultures, and also claim favorable results. Mathes,70 continuing the work of Gottstein after the death of the latter, employed the method of immunizing with the product ob- tained by digesting typhoid bacilli with trypsin. The poison so pro- duced he speaks of as "fermotoxin." Liidke 71 slightly modified the Gottstein-Mathes method by digesting the typhoid bacilli with pepsin and hydrochloric acid, and with the poison so produced immunized 8 goats, reenforcing the immunization by the subsequent injection of the bacilli themselves. He claims that 0.05 to 0.1 c. c. of the serum so produced protected animals against five times the lethal dose of the poison. In a small series of human cases treated by this method he reports good results. Garbat and Meyer 72 immunized animals with sensitized typhoid bacilli, and claim that the most potent sera for typhoid immunization can be obtained by the combination of sera produced by the injection of sensitized and of unsensitized bacteria. They assert that the typhoid bacillus contains two definite antigens, one particularly as- 64 Kraus and Stenitzer. Wien. kl. Woch., Vol. 20, 1907, pp. 344 and 753, and Vol. 21, 1908, p. 645. 65 MacFadyen. Cited from Stenitzer in "Kraus und Levaditi Handbuch," Vol. 2. 66 Arima. Centralbl. f. Bakt., 65, 1912, p. 183. Orig. 67 Kraus and Stenitzer. Wien. kl. Woch., Vol. 22, 1909, p. 1395; Deutsche med. Woch., March, 1911. 68 Herz. Wien. kl. Woch., Vol. 22, 1909, p. 1746. 69 Rodet and Lagrifoul. C. R. de la Soc. de Biol, April, 1910. 70 Mathes. D. Archiv f. kl. Med., Vol. 95, 1909. 71 Liidke. D. Archiv f. kl. Med., 98, 1910. 72 Garbat and Meyer. Zeitschr. f. exp. Path. u. Ther., Vol. 8, 1911. 478 INFECTION AND RESISTANCE sociated with the bacterial ectoplasm, which becomes active when the bacteria enter the animal body, and a truly endocellular poison which does not become active until the surrounding ectoplasm is dissolved. They believe that sensitizing bacteria is a method for the production of endotoxin, and think that therefore the ideal serum for the treat- ment of typhoid consists of a mixture of two sera produced each with one of the antigens, that is, with sensitized and unsensitized bacteria. Rommel and Herman 73 did not obtain encouraging results with this serum. From a study of the literature it seems to us that in spite of the many different methods of production employed by various observers in their studies on typhoid sera it is quite likely that all these sera are essentially alike, containing, quantitatively, according to the de- gree of immunization, bactericidal, agglutinating, and opsonic prop- erties, with possibly a limited amount of neutralizing power for the poisons liberated from the typhoid bacilli in the body. As far as we can judge from clinical reports the therapeutic value of the sera so far produced is not very great. It seems that cases treated early in the disease may be benefited, and possibly an early cessation of the bacteriemia can in this way be attained. However, it does not seem either theoretically or from the study of clinical publications that any very marked effects have followed the use of any of the sera in advanced cases. THE SEKUM TREATMENT OF PLAGUE That the serum of animals immunized with killed plague cultures may actively protect normal animals from experimental infection was first shown by Yersin, Calmette, and Borrel.74 The serum which they produced possessed apparently powerful bactericidal action, but no antitoxic properties were demonstrated. They determined its protective powers by injecting measured quantities into mice and infecting them with fatal doses of virulent plague bacilli 24 hours later. The Yersin serum which was produced for the treatment of plague as a result of these experiments was made, then, by the gradual immunization of horses with first dead plague bacilli, finally with virulent living organisms. The serum has been extensively used by many observers with results that leave one much in doubt as to its efficacy. Yersin 75 himself, reporting on an epidemic in IS^hatrang, reports a general mortality of 73 per cent, for the whole epidemic, a mortality of 100 per cent, in untreated cases, and of 42 per cent, among those treated with his serum. Good results were also reported from the epidemics in Amoy and Canton in 1896.. 73 Rommel and Herman. Centralbl. f. Bakt. Ref . Vol. 53, 1912. 74 Yersin, Calmette, and Borrel. Ann. de I'Inst. Past., 1895. 75 Yersin. Ann. de I'Inst. Past., 1899. THERAPEUTIC IMMUNIZATION IN MAN 479 However, these results apparently were not accepted by all observers as proving the efficiency of the serum, since the number of cases observed were few, and the irregularity in the gravity of the disease in different individuals makes statistical evidence unreliable unless large material can be studied. Kolle and Martini 76 announce that Dr. Choksy reported very poor success with the. Yersin serum, and cite a number of later writers whose results with this serum were also unsatisfactory when used on human beings. That the serum unquestionably contains antibodies against the plague bacillus is | testified to, not only by the French observers themselves, but also by the German Plague Commission of 1899, and by Kolle and Mar- tini 77 themselves. The Commission experimented with this serum upon monkeys, and showed that it possessed unquestionable protec- tive powers in rodents and in monkeys when given 24 hours before the plague infection, and in monkeys possessed fair curative prop- erties when injected 24 hours later than inoculation with the plague bacilli. Because of the doubtful success in the treatment of human beings with this serum Yersin and Roux at the Pasteur Institute later altered their methods of serum production by injecting, not only dead and living plague cultures, but considerable quantities of culture filtrates after the horses had attained a high degree of im- munity. Later observations on the Yersin 78 serum have been pub- lished by the British. Plague Commission in 1908 and 1911. In this investigation the cases were controlled as to their severity by blood culture, since it had been claimed by a number of earlier investi- gators that the Yersin serum was efficient in mild cases, but failed entirely in the severe ones. It seems from the report of this Commis- sion that ordinarily 70 per cent, of cases of plague without bacilli in the blood survive while three-quarters of those with mild septicemia die, and all of those with a marked septicemia succumb. In the summary given of 146 cases treated with Yersin' s serum by the British Commission 65.1 per cent, died, whereas of 146 untreated controls 71.90 per cent. died. These figures, together with an analysis of the percentages, classified according to the severity of the infections, do not show a very marked curative action on the part of the serum. Markl,79 who claims that the plague bacillus produces a soluble toxin, has produced a plague serum by immunization of animals by filtrates of broth cultures. He claims that 0.1 c. c. of his serum, as produced at Vienna, will protect various animals against lethal doses of plague bacilli if given at the same time. He attributes much of 76 Kolle and Martini. Deutsche med. Woch., 1902, p. 29. 77 German Plague Commission. Arb. a. d. kais. Ami., Vol. 16, 1899. 78 British Plague Commission. Joufn. of Hyg., Vol. 12, Sup., 1912, p. 326. 79 Markl. Centralbl. f. Bakt., 24, 1898; Zeitschr. f. Hyg.. 37, 1901; Zeitschr. f. Hyg., 42, 1903. 480 INFECTION AND RESISTANCE its curative action to the fact that in the presence of this serum active phagocytosis takes place. Dean,80 in 1906, also claimed to have produced strongly antitoxic plague sera by treating horses with filtrates from 8 to 10 weeks old bouillon cultures. He claims that 1 c. c. of his serum will neutralize 150 or 450 minimal lethal doses of the plague poison according to whether one measures the M L D by death in 48 hours or in 4 days; Rowland 81 also has produced a serum by the immunization of ani- mals with the "toxins" produced by his sulphate process. Rowland has apparently utilized the idea previously advanced by Lustig of immunizing with "nucleoproteins" derived from the plague bacillus instead of with the whole bacteria. Lustig's 82 method consisted of washing up agar cultures of plague bacilli in 1 per cent, sodium hydrate solution, precipitating with ascitic acid, taking up the pre- cipitate in an indifferent fluid and injecting it into horses. The serum produced by Lustig's method was used in Bombay, and is re- ported by Hahn as effective in milder cases, but without action in the more severe ones. There was but slight difference in the latter type between the treated and the untreated cases. Rowland's 83 method consisted in the treatment of the moist bacteria with enough anhydrous sulphate of soda to combine with all the water present, freezing and thawing the mixture and filtering off the bacterial deposit at 37° C. Subsequently he extracted this bacterial mass with water. The extract so obtained was fatal to rats in quantities of 0.05 to 0.1 mg., killing them in 18 hours. In his experiments doses of 0.001 to 0.01 afforded protection, the last- named quantity reducing the mortality after inoculation of fatal doses from 80 per cent, to 10 per cent. The sera produced by the immunization of horses with these supposed nucleoproteids are taken to be antitoxic in nature by Row- land himself and by MacKonky. They were used in the treatment of plague cases in the epidemics of 1908 and 1911 by the Maratha Hos- pital in Bombay, and reported upon by the Indian Plague Commis- sion on the basis of observations made by Dr. Choksy. The cases in this series were controlled, as were those treated by the Yersin serum, by blood culture. Here the results were not striking — 68.40 per cent, of the serum-treated cases died, while 77.60 per cent, of the controls died. Altogether we cannot draw any definite conclusions as to the value of the serum treatment in plague. On the whole it does appear 80 Dean. Cited from MaeKonky, Journ. of Hyg., Vol. 12, Plague Suppl. II, 1912, p. 402. 81 Rowland. Journ. of Hyg., Vol. 11, Plague Suppl. I, pp. 11-19. 82 Lustig1. "Monograph Sierotrapia e Vaccin Prev. Control la Peste," Turin, 1899 ; cited from Kolle and Martini, loc. cit. 83 Rowland. Journ. of Hyg., Vol. 10, p. 536. THERAPEUTIC IMMUNIZATION IN MAN 481 that the milder cases are materially benefited by the treatment, and it is not at all impossible that in such cases aggravation of a milder case into fatal septicemia may be prevented by the timely adminis- tration of the plague serum. Animal experimentation also seems to indicate that the administration of the serum may be of great value as a prophylactic measure. It seems, on the other hand, as far as we can judge from the evidence of statistics, that when a case of plague has developed into the condition of active septicemia the administra- tion of even the strongest plague sera at present available is of little use. And this is indeed unfortunately true of all passive immuniza- tion where the activity of the serum seems to depend chiefly upon bactericidal and opsonic properties. For we cannot definitely accept at the present day the claims that a true antitoxic serum, in the sense of those produced against diphtheria and tetanus poisons, can be really produced in the case of plague. The toxic substances derived from plague bacilli by a number of observers do not correspond in many particulars to true toxins. FACTS CONCERNING ACTIVE PROPHYLACTIC BOCTJNIZATION IN MAN In a previous chapter we have dealt with the treatment of in- fectious disease with emulsions of dead bacteria or vaccines. The discussion there was confined to the use of these substances in the case of developed disease in which the infectious agent had already gained a foothold in the body. Concerning this form of therapy much difference of opinion exists, and we have seen that careful judgment must be applied to the selection of cases to which treatment with vaccines is adapted. Concerning the prophylactic immunization of human beings with bacteria there can be little difference of opinion ; this procedure finds its justification in prolonged laboratory experience in the hands of many men since the days of Pasteur. The principle of specifically increasing the resistance of an in- dividual by treatment with an altered form of the disease, either with the attenuated bacteria, with dead bacteria, or with bacterial extracts, has been sufficiently discussed in Chapter IV. It is indeed surprising that this phenomenon of prophylactic protection, though discovered by Jenner in small-pox, and developed by Pasteur in rabies, did not find more general application to the diseases of man until recent years. At present such methods are in extensive use in typhoid fever, in which they have had unquestionably excellent re- sults. In the cases of cholera and plague numerous attempts have been made, but the results here are not as clear-cut as they have been in the case of typhoid. In the succeeding paragraphs we have set 482 INFECTION AND RESISTANCE forth as briefly as possible the methods employed in prophylactically immunizing man against this disease in which this procedure has been most commonly attempted. PROPHYLACTIC IMMUNIZATION IN TYPHOID FEVER The first attempt to inoculate human beings with typhoid bacilli with the intention of producing prophylactic active immunization was probably that made by Pfeiffer and Kolle 84 in 1896. During the same year also Wright 85 made similar studies in England, and soon after this he reported upon the development of antibodies in the blood of 17 people inoculated with typhoid. By these studies it was shown that human beings could be inoculated with dead typhoid bacilli without danger, and this logically led to the attempt to vac- cinate human beings on a large scale. It is hardly necessary to dwell upon the desirability of such a procedure. From tables recently published by Russell 86 we take the information that, in our own Spanish-American war, 20,738 cases of typhoid with 1,580 deaths occurred in a total enlistment of 107,- 973. In this entire war 243 men were killed in action or died of their wounds, while almost 7 times as many died of typhoid fever. In the British army during the Boer war there were over 75,000 cases of typhoid in 380,000 men, and in the Russian army during the Russo-Japanese war over 17,000 cases of typhoid occurred, over half as many as the number of men killed in action. Such appalling fig- ures leave no possible doubt as to the desirability of prophylactic im- munization in armies, and there can be little question that typhoid fever is sufficiently prevalent in many parts of the civilized world to encourage prophylactic immunization of individuals, even when not living under the especially dangerous conditions of camps. Following the preliminary studies of Pfeiffer and Kolle and of Wright extensive practical studies of vaccination were made in the German colonial army during the Herrero war, and by British bacteriologists during the Boer war. Leishmann 87 also studied care- fully the results of vaccination among regiments of the British army in India. The vaccine employed by Wright and his associates in England consisted of broth cultures of a typhoid bacillus killed by exposure to 53° C., and by the further addition of 0.4 per cent, of lysol. The German vaccine consisted of emulsified agar cultures similarly killed. The results obtained with these vaccines, although encouraging, 84 Pfeiffer and Kolle. Deutsche med. Woch., 22, 1896, p. 735. 85 Wright. Brit. Med. J., Jan., 1897, p. 256. 86 Russell. Amer. J. of Med. Sciences, Dec., 1913, Vol. 146. 87 Leishmann. Glasgow Med. Journ., 1912, Vol. 77, p. 408, cited from Russell. THERAPEUTIC IMMUNIZATION IN MAN 483 were not as striking as had been hoped. Russell 88 summarizes the earlier attempts by stating that the morbidity was reduced to about one half among vaccinated persons with a slightly greater reduction of mortality. The last-named writer also attributes the early fail- ures to the overheating of the vaccines with a consequent reduction of their antigenic properties, and to timidity in their administration resulting from Wright's fear of a negative phase. Russell, of the United States Army Medical Corps, made a most extensive study of typhoid vaccination in this country. After careful consideration of the methods of others he produces his vaccines as follows : A single strain of typhoid bacilli is used (culture Rawlings obtained from England), and this is grown on agar in Kolle flasks for 18 hours. The purity of the culture is tested out both morphologically and by transplantation upon the double sugar media devised by Russell. Agglutination tests are also made. After 18 hours the growth is washed off in small quantities of salt solution and the emulsion heated in a water bath for one hour at 53° C. ; it is then diluted with sterile salt solution to a concentration of one billion bacteria to a cubic centimeter. Then 0.25 per cent, of tricresol is added. Before use the safety of the vaccine is ascertained both by aerobic and anaerobic cultivation and by the injection into mice and guinea pigs of consid- erable quantities for the exclusion of possible tetanus contamination. The efficiency of the vaccine is then tested by injecting rabbits with three doses at 10-day intervals, and determining the resulting ag- glutinating power. With these vaccines under the direction of the United States Army Medical Corps the troops mobilized in Texas, California, and along the Mexican border were treated. Compulsory vaccination was established in March, 1911, and the results as reported by Russell have fully justified the measure. The following table taken from Russell's paper will illustrate the results obtained : Typhoid Fever. Officers and Enlisted Men, United States Army Totals Yr. Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. for 9 months Volun- ( 1908 5 6 4 2 3 11 14 31 25 26 12 8 101 tary (1909 4 10 6 4 11 15 26 14 16 45 20 6 106 (1910 8 11 1 4 2 6 12 27 21 16 20 11 92 1911 3 3 3 7 4 4 4 7 4 4 1 0 39 Com- pulsory 1912 1 2 2 0 0 3 1 3 1 4 0 1 13 1913 0 0 0 0 0 0 0 0 - 0 0 0 0 0 Paratyphoid fever included in figures for 1908, but excluded in other years. Cases paratyphoid, 1909, 3; 1910, 3; 1911, 2; 1912, 3; 1913, 0. 88 Russell. Am. Journ. of Med. Sc., Vol. 146, Dec., 1913. 484 INFECTION AND RESISTANCE We have mentioned in another place that Metchnikoff and Bes- redka in their studies on typhoid vaccination in the chimpanzee have concluded that very little protective value resided in vaccina- tion with dead typhoid vaccines, whereas animals vaccinated with small amounts of living cultures were very efficiently protected. Metchnikoff and Besredka adopted finally the method of immunizing with living sensitized vaccines. By this is meant typhoid bacilli that have been exposed to the action of heated immune serum, or, in other words, typhoid bacilli that have absorbed specific antibodies. There is no question as to the efficiency of this form of vaccination. The method of employing sensitized bacteria for these purposes utilized by Besredka in the case of plague has unquestionably won an important place in active immunization. However, the results of Eussell and others seem to indicate that in human beings the use of dead vaccines is certainly of considerable value, and there are certain practical objections to the use of living vaccines in immuniza- tion of large numbers of people as in armies to which Russell calls attention. In the first place, living vaccines cannot be stored for any considerable period, and may become a source of possible infec- tion by mouth if carelessly handled; furthermore, contamination is not so easily ruled out in the case of living vaccines when used on a large scale, and it is not possible at present to require compulsory vaccination with living bacteria. Gay has recently recommended the use of sensitized killed vac- cines. He controls the efficiency of his vaccines by testing them out on rabbits in which typhoid septicemia has been produced by inocu- lation with cultures grown on rabbit blood agar. These vaccines have not yet been used upon sufficient numbers to justify conclusions. It would seem, however, that any one of the methods mentioned must possess considerable value, since they all represent merely slight variations of the same procedure. The method at present used in the German, British, and American armies, namely, vaccination with dead cultures, seems certainly, according to RusselPs statistical studies, to have yielded excellent results and recommends itself by its extreme simplicity and safety. ACTIVE PROPHYLACTIC IMMUNIZATION IN CHOLERA Attempts to protect human beings against cholera by prophylactic vaccination were made as early as 1885 by Ferran,89 a pupil of Pasteur. At the time at which Ferran' s experiments were done little was known regarding the production of immunity with killed cul- tures or with bacterial extracts, and Ferran, under the influence of the French school and its endeavors to immunize with living attenu- 89 Ferran. C. E. de VAcad. des Sc., 1885. THERAPEUTIC IMMUNIZATION IN MAN 485 ated organisms, applied similar" methods to cholera. First experi- menting with guinea pigs, he soon applied his method to human beings, inoculating them with small quantities of living broth cul- tures of cholera spirilla. In many of his experiments he gave, at the first injection, 8 drops of a fresh broth culture, following this after 8 days with 0.5 c. c. of a similar culture. There is no reason why Ferran's method should not have yielded excellent results. How- ever, it is stated that he worked with impure cultures, and other observers, notably Mkati and Eietsch, van Ermengen, da Lara, and others, failed to obtain encouragement in their subsequent investiga- tion of this method of vaccination. The method which Haffkine 90 worked out some years after Fer- ran's experiments also depended upon the injection of living cul- tures, but Haffkine attempted, by a rather elaborate technique, to produce two separate vaccines, one attenuated, the other enhanced in virulence. Attenuation was accomplished by growing the cholera spirilla at a temperature of 39° C. in broth over the surface of which a constant stream of sterile air was passed. Under these conditions the first crop of cholera organisms died rapidly, but Haffkine prac- ticed reinoculation into new broth flasks before complete death of the original culture had taken place; after a series of generations of cultivation in this way he obtained cultures which produced merely temporary and slight edema when injected under the skin of guinea pigs. This weakened virus was used for the first inoculation. He enhanced the virulence of cholera cultures with the purpose of producing a strain of maximum potency comparable to virus fixe. His procedure was as follows : a. Giving an animal an intraperitoneal injection of cholera spirilla larger than the fatal dose. b. Taking out the peritoneal exudate and exposing it for a few hours to the air. c. Injecting this exudate into another animal and treating the resulting peritoneal exudate in the same way. After a series of such animal passages he claims to have obtained a virus of great virulence, and this is his second and stronger vaccine. In applying the method to human beings Haffkine planted the cholera spirilla upon agar slants of the standard size, emulsified the growths in sterile water, and injected 1/5 to 1/20 c. c. of such a cul- ture, using first the weak vaccine and five days later a more virulent culture. Beginning his work as early as 1893, Haffkine and others vac- cinated as many as 40,000 people in India. On the whole, the results obtained were very encouraging. It is a question, however, whether or not his method is unnecessarily complicated. In the light of our 90 Haffkine. The Lancet, February, 1893; Brit. Med. Journ., December, 1895. 486 INFECTION AND RESISTANCE more recent knowledge concerning cholera immunity it seems likely that the importance which Haffkine attached to the virulence of the cholera culture used for injection was exaggerated, and we have reason to believe that simple immunization with killed cultures may produce results fully as efficacious. After all, we could not expect, at least at present, to produce by active artificial immunization an immunity as permanent as that which results from an attack of the disease. Concerning the reasons for the acquisition of such perma- nent immunity we have as yet little knowledge. Even Haffkine's method of inoculation with living virus does not, by his own estima- tion, last longer than possibly two years. It is therefore likely that prophylactic immunization in cholera is efficacious by reason of the appearance in the blood serum of the specific bactericidal and opsonic substances by which the small numbers of cholera organisms entering during spontaneous infection can be disposed of before a foothold in the body is gained. Tamancheff later used Haffkine's method, but killed the cultures by the addition of a 0.5 per cent, solution of carbolic acid. Kolle 91 later recommended the injection of dead cholera or- ganisms, maintaining that a single injection of about 2 milligrams of a culture killed by exposure to 50° C. for a few minutes, and by the addition of 0.5 per cent, of phenol, is sufficient to immunize suc- cessfully. Good results with Kolle' s method have been reported from Japan. Strong,92 also proceeding from the idea that the immunizing antigen is present, as such, within the cell body of the cholera spirilla, recommends the injection of autolytic products obtained by digesting cholera spirilla in aqueous suspension and filtering. He prepared his aprophylactic" by growing the organisms upon agar, then suspending the growth in sterile water and keeping it at 60° C. for from one to twenty-four hours. The mixture was then exposed to 37° C. for from two to five days and filtered through Reichel filters. One to 5 c. c. of this was used in his experiments upon human beings. PKOPHYLACTIC IMMUNIZATION AGAINST PLAGUE The first attempts to immunize human beings prophylactically against plague were those of Haffkine.93 The first vaccinations were carried out with broth cultures killed at 65° C. He tested out his vaccines on a large scale in Bombay, and obtained apparently prom- ising results. In a plague epidemic occurring in a Bombay prison 91 Kolle. Deutsche med. Woch., 1897, No. 1. 92 Strong. Journ. Inf. Dis., Vol. 2, 1905. 98 Haffkine. Bull, de I'Inst. Past., Vol. 4, 1906, No. 20, p. 825. THERAPEUTIC IMMUNIZATION IN MAN 487 only 2 of 151 vaccinated persons became ill, and neither of these died; whereas, of 177 unvaccinated persons 12 became ill and 6 died. In large series of vaccinated people only 1.8 per cent, were infected with plague, with a mortality of 0.4 per cent, for the total, whereas of unvaccinated individuals in the same epidemic 7.7 per cent, fell victim to the disease, with a mortality of 4.7 per cent. The German Plague Commission, consisting of Gaffky, Pfeiffer, and Dieudonne, recommended a vaccine of killed agar cultures. Kolle and Otto,94 basing their earlier results upon experiments car- ried out with monkeys, mice, guinea pigs, and rats, have come to the conclusion that vaccination with dead plague cultures is much in- ferior to that obtained when attenuated living cultures are used. The same conclusion has been reached by Kolle and Strong.95 Kolle and Otto found that the immunization of animals with large doses of killed agar cultures of plague bacilli and with Haffkine's prophy- lactic did not protect them against subsequent inoculation with virulent cultures. Strong 96 subsequently made a very careful comparative study of the various methods of plague vaccination, and concluded that the most efficient method is immunization with attenuated living cul- tures. He showed that when carefully done this method can be safely employed in human beings, but admits that his work must be as yet considered as experimental and further studied before it can be universally employed. Besredka 9T has advised the use of sensitized dead plague cul- tures, claiming, from animal experimentation, that such vaccines produce an efficient and relatively durable immunity. Rowland 98 confirms the immunizing properties of Besredka's vaccines in plague, and believes that the antigenic properties of the plague bacillus are attached to the bacterial nucleoproteins, and can be extracted with these. Rowland prepares a vaccine by the treat- ment of the moist bacteria with enough anhydrous sodium sulphate to combine with all the water present, freezing and thawing the mixtures, then filtering off the bacterial deposits at 37° C., and ex- tracting them with water. The solution so obtained was fatal to rats in small quantities and afforded substantial protection, reducing the mortality on subsequent inoculation of a standard culture from 80 to 10 per cent. 94 Kolle and Otto. Deutsche med. Wocli., 1903, p. 493, and Zeitschr. f. Hyg., Vol. 45, 1903. 95 Kolle and Strong. Deutsche med. Woch., XXXII, 1906, p. 413. 96 Strong. Journ. of Med. Res., N. S., 13, 1908. 97 Besredka. Bull de I'Inst. Past.. Vol. 8, 1910. 98 Rowland. Journ. of Hyg., VoL 12, 1912, p. 344. 488 INFECTION AND RESISTANCE PROPHYLAXIS AGAINST SMALL-POX In the case of small-pox the general method of active prophylactic immunization is in principle identical with that devised by Jenner in the 18th century. The original observation from which Jenner worked was that dairy maids and other individuals who had been infected with cow pox were thereafter spared when a small-pox epi- demic appeared in the region in which they lived. It is now agreed by most observers who have studied the problem that the virus of cow pox and that of small-pox are identical in nature ; the former repre- senting a strain attenuated by passage through the animal body. This is based chiefly upon the observation that true variola can be transmitted to cattle, and that it can be thus carried from animal to animal, during this process becoming attenuated for human beings to such a degree that reinoculated into man a simple vaccinia is produced." Small-pox, therefore, represents in principle active immunization by means of attenuated virus. When vaccination was first introduced the virus was taken from preceding pustules produced in other human beings. This has been given up in most countries to-day largely be- cause of the dangers of transferring syphilis and other diseases in this way. At present the method of obtaining virus for vaccination purposes is carried out as follows: The initial material consists of what is known as "seed" virus, which can be obtained from spon- taneous cow pox or from vaccination pustules in children, or again from pustules obtained in calves after several passages of small-pox virus through these calves. From such seed virus calves may be in- oculated for vaccine production or else the calves may be inoculated from the material obtained from other calves in the usual way. Healthy young animals are used; they are washed along the abdomen, strapped down upon specially prepared tables, and the abdominal skin thoroughly cleansed with soap and water. The exact procedure varies in different places; often the skin is thor- oughly cleansed with carbolic solution, and this is thoroughly re- moved with sterile water before inoculation, or else cleansing is relied upon without the use of germicides. Over the clean area longitudinal scratches 1 to 2 c. c. apart are made, and into these the seed virus is rubbed. The animals are then kept in a clean stall, preferably over asphalt floors, and rigid cleanliness is observed during the period of development of the pustules. After the 6th or 7th day, when the vesicles are beginning to appear, the abdomen is well washed and cleansed of superficial dirt without the use of an antiseptic, and the pulp removed from the lesions with a curette. The pulp so removed is placed into 60 per cent, glycerin and thor- oughly ground up in a specially constructed mill. According to Rosenau, the animal should always be killed before the vesicles are- 99 Haccius. Cited from Paul, Kraus and Levaditi, Vol. 1, p. 593. THERAPEUTIC IMMUNIZATION IN MAN 489 removed, not only for humane reasons, since the same object might be attained with anesthesia, but because a thorough autopsy can then be performed to determine the health of the calf. Vaccines so obtained always contain bacteria, the glycerin there- fore serving a double purpose : one, the preservation of the virus, the other a gradual destruction of the bacteria. Kosenau has shown that the addition of 2 to 4 parts of 60 per cent, glycerin to one part weight of the pulp prevents the growth of bacteria and probably destroys them by dehydration. Most of the bacteria are destroyed within one month at 20° C. During this period, then, from 4 to 6 weeks, the glycerinated virus should not be used, and should from time to time be controlled by cultivation. At the end of this time the lymph is ready for use. Formerly the material for the vaccination of human beings was obtained very simply by dipping ivory splinters into the fluid of pus- tules, allowing this to dry, and rubbing these ivory or bone points into the exudate obtained by scratching the skin of the individual to be vaccinated. This method has practically gone out of use, and to-day the ripened glycerin pulp prepared as above is taken up in small capillary glass tubes and from these blown upon the vaccina- tion scratch. The efficiency of vaccine virus can be tested for po- tency by the inoculation of the ears of rabbits before use. ACTIVE PROPHYLACTIC IMMUNIZATION IN RABIES (HYDROPHOBIA) Although many modifications have been suggested and actually used in different parts of the world, the most common method of immunizing against rabies still remains that originally devised by Pasteur. The Pasteur treatment takes advantage of the prolonged incubation period of rabies and is planned to confer immunity be- tween the time of inoculation and the time at which the disease would naturally appear. Since this period in ordinary street infec- tion by dog bite is usually 40 days or more, a considerable interval for active immunization is available. Formerly much of this time was lost in that the diagnosis of hydrophobia in the dog or other animal that had caused the injury could not be made with certainty until the results of rabbit inoculations had been obtained. Nowadays the ease with which a diagnosis of hydrophobia can be made within a few minutes by finding negri bodies in the hippocampal and cere- bellar cells has added considerably to safety in that it has made pos- sible a gain of almost two weeks in determining whether treatment should be instituted or not. Here again, although the infectious agent of rabies is not known with certainty even at the present day, the method of Pasteur de- pends upon active immunization by means of an attenuated virus. In standardizing the virus for the purpose of treatment Pasteur first produced what he calls the "virus fixe." This consists of the 490 INFECTION AND RESISTANCE ordinary street virus as obtained from rabid animals passed through a considerable series of rabbits (20-30) until its virulence for these animals has reached a maximum. After a sufficient number of such rabbit passages the incubation time after intracerebral inoculation is reduced to 7 or 8 days, but can no longer be shortened by further passage. The brain and cord material of rabbits dead of rabies after such repeated passages constitutes virus fixe. This can be preserved for considerable periods in 60 per cent, glycerin, and this is the initial material from which the attenuated preparations for treat- ment are produced. In preparing the material for treatment a small amount of virus fixe is injected subdurally into rabbits, about 0.2 c. c. of a salt solu- tion emulsion being given. The inoculation is very easily made through a small trephine opening in the skull, and contamination is very easily avoided. Just before the rabbit dies when completely paralyzed it is killed by chloroform and the cord is removed best by the method of Oschida.100 The rabbit is nailed to a board, back up- permost, and washed with a weak antiseptic, a longitudinal incision is then made along the backbone from the occiput to the lumbar region, and the vertebral column laid bare. After searing the tissues around the back of the head the spine is cut across just behind the occiput, and again in the same way just above the sacrum. The neck and lumbar regions are dissected loose from the skin and gauze is inserted under it to avoid contamination. The assistant grasps the end of the spinal cord as it appears in the cervical region and pulls on it very slightly while the operator with a glass rod or a piece of wire pushes against it from below. If this is carefully done the spinal nerves are torn and the cord can be gradually pulled out of place. This procedure is by far the best, although it requires a certain amount of practice. The cords so removed are hung up by a thread in bottles contain- ing sticks of caustic potash and exposed in a dark place to 22° to 23° C. Under these conditions of drying and temperature the virus is gradually attenuated until at the end of 13 days or more the viru- lence is practically nil. If removed from the drying bottles at any time during the process and kept in a refrigerator in sterile glycerin the virulence, whatever it may be at the time of placing into the glycerin, remains fairly constant for long periods. When any of this material is used for treatment little pieces of the cord % cm- in length are cut off and emulsified in 2.5 c. c. of salt solution, and this emulsion is used for injection.101 100 Oschida. CentraXbl. f. Bakt., Vol. 29, 1901. 101 In our description of the methods of drying1 rabies, for the sake of adhering to a standard, we follow closely the directions laid down by A. M. Stimson, in the U. S. P. H. S. Bull. 65, 1910. There are various modifica- tions used in different countries, in many cases unimportant, and it seems well to adhere to the U. S. regulations as a standard for this country. THERAPEUTIC IMMUNIZATION IN MAN 491 When patients are to be treated the principle of the treatment is to inoculate them first with cords that have been dried for consider- able periods, gradually proceeding toward those that have been dried for less prolonged times and are therefore more virulent. The treat- ment is varied in the individual case according to the severity of the injury. Formerly treatment was begun with cords dried as long as 16 days. More recently it has been found that cords dried for longer than 8 days are practically non-virulent and correspondingly lack in antigenic value. They are no longer employed therefore, since their use is regarded as a waste of time. The following tables taken from Stimson's article in Bulletin 65 of the Hygienic Laboratory of the U. S. Public Health Service give the standard methods of treat- ment as recommended by the United States Public Health Service: Scheme for Mild Treatment Amount injected Amount injected Cord Cord Day (injections) Adult 5-10 1-5 Day (injections) Adult 5-10 1-5 (c. c.) yrs. (c. c.) yrs. (c. c.) (c. c.) yrs. (c. c.) yrs. (c. c.) ' 1 8-7-6 = 3 2.5 2.5 2.0 12 4=1 2.5 2.5 2.5 2 5-4 = 2 2.5 2.5 1.5 13 4 = 1 2.5 2.5 2.5 3 4-3 = 2 2.5 2.5 2.0 14 3 = 1 2.5 2.5 2.0 4 5=1 2.5 2.5 2.5 15 3 = 1 2.5 2.5 2.0 5 4=1 2.5 2.5 2.5 16 2 = 1 2.5 2.0 1.5 6 3 = 1 2.5 2.5 2.0 17 2 = 1 2.5 2.0 1.5 7 3 = 1 2.5 2.5 2.0 18 4=1 2.5 2.5 2.5 8 2 = 1 2.5 1.5 1.0 19 3 = 1 2.5 2.5 2.5 9 2 = 1 2.5 2.0 1.5 20 2 = 1 2.5 2.5 2.0 10 5 = 1 2.5 2.5 2.5 21 2=1 2.5 2.5 2.0 11 5 = 1 2.5 2.5 2.5 Scheme for Intensive Treatment Amount injected Amount injected Cord Cord | Day (injections) 5-10 1-5 Day (injections) 5-10 1-5 Adult yrs. yrs. Adulf yrs. yrs. (c. c.) (c. c.) (c. c.) (c. c.) (c. c.) (c.c.) 1 8-7-6 = 3 2.5 2.5 2.5 12 3 = 1 2.5 2.5 2.0 2 4-3 = 2 2.5 2.5 2.0 13 3 = 1 2.5 2.5 2.0 3 5-4 = 2 2.5 2.5 2.5 14 2 = 2.5 2.5 2.0 4 3=1 2.5 2.5 2.0 15 2 = 2.5 2.5 2.0 5 3 = 1 2.5 2.5 2.0 16 4 = 2.5 2.5 2.5 6 2=1 2.5 2.0 1.5 17 3 = 2.5 2.5 2.5 7 2 = 1 2.5 2.5 2.0 18 2 = 2.5 2.5 2.0 8 1 = 1 2.5 1.5 1.0 19 3 = 2.5 2.5 2.0 9 5 = 1 2.5 2.5 2.5 20 2 = 2.5 2.5 2.5 10 4=1 2.5 2.5 2.5 21 1 = 1 2.5 2.5 2.0 11 4=1 2.5 2.5 2.5 492 INFECTION AND RESISTANCE This is the standard treatment used almost everywhere in the world at present. Other methods have been recommended. One of these is that of Hogyes, in which virus fixe unattenuated is used in dilution. Hogyes begins by injecting 3 c. c. of a 1 to 10,000 dilution of virus fixe, gradually proceeding within 14 days to 1 c. c. of a 1 to 100 dilution. Fixed virus attenuated by the addition of antirabic serum and chemical disinfectants (carbolic acid) and by partial digestion in gastric juice has also been used, but none of* these methods has at- tained widespread application. CHAPTER XX ABDEKHALDEN'S WOEK UPON PEOTECTIVE FEK- ME1STTS OF THE ANIMAL BODY THE recent researches of Abderhalden 1 upon the intravascular digestion of foreign substances introduced into animal bodies promise to have considerable bearing upon problems of immunity. Abder- halden, whose work we cite chiefly from his monograph, "Die Schiitz- fermente des tierischen Organismus," took as his point of departure the conception that the animal body must necessarily dispose over a mechanism whereby it can assimilate foreign substances which ob- tain entrance unchanged into the circulation. In our section upon the nature of the precipitins, especially in the discussion of Gengou's conception of "albuminolysins," we have called attention to the probable significance of protein antibodies as a mechanism for the disposal of such foreign substances. In the bodies of the higher animals in which a special alimentary system, with its many diges- tive ferments, is well developed, it is most probable that the normal condition of digestion is one in which the foreign substances utilized for nutrition are completely split into their simpler components be- fore they gain entrance to the circulation. Nevertheless, abnormal conditions or accidents, such as gastro-enteric diseases, digestive dis- turbances, and bacterial infections, may lead to a condition, prob- ably frequent enough in ordinary life, during which such foreign substances may get into the blood stream without previous cleavage. The problem is to determine where and how such substances, protein or otherwise, are broken up so that they may be either assimilated or eliminated. We have referred in another place to the fact that foreign proteins may occasionally pass through the kidneys and be eliminated unchanged. This has been shown actually to occur by Oppenheimer, Ascoli, and others, but probably represents a very unusual state of affairs produced by special experimental conditions. As a rule these substances are disposed of within the body by chemi- cal cleavage or by assimilation. Abderhalden believes that this process depends upon the mobilization of "protective ferments," a term which he borrows from Heilner,2 and suggests the possibility 1 Abderhalden. "Schiitzfermente des tierischen Organismus," Springer, Berlin, 1912. 2 Heilner. Cited from Abderhalden Zeitschr. f. BioL, Vol. 50, 1907. 493 494 INFECTION AND RESISTANCE that these ferments may possibly originate in the leukocytes. He re- fers to the work of Friedrich Miiller, in which it was shown that the resorption of pneumonic consolidations is largely carried on by leuko- cytic ferments. Moreover, we possess in support of such a conception the many consistent reports of the successful extraction of various ferments from leukocytes, some of which are referred to in detail in another section. Experimentally Abderhalden approaches his problem by deter- mining the presence of specific ferments in the blood of animals into which various foreign substances have been introduced by paths other than the alimentary canal. For this purpose he has developed a number of methods, the most important of which are his optical method and his dialysis method. The optical method used for the determination of the proteolytic properties of the serum depends upon the fact that many of the amino-acids are optically active. Moreover, most of these substances are chemically known and their optical activity determined, so that it is possible to take blood serum which is to be examined for its contents of particular ferments, mix them with a suitable protein, or preferably a polypeptid, and de- termine with a polariscope the rotation which takes place. We will not go into the technique of this method more extensively because we have no personal experience with it, and the method is one of such delicacy that it is best obtained from Abderhalden's original publications directly.3 His dialysis methods depend upon placing the blood serum and fermentable substance into dialyzing bags, suspend- ing them into distilled water, and determining the presence of pep- tone, amino-acids, or total nitrogen in the liquid outside of the bag after definite intervals of time. By these and other methods Abderhalden 4 has carried out tests with a large number of different substances. Experimenting first with proteins, he injected egg albumen, horse serum, silk peptone, gelatin, edestin, casein, etc., into dogs and rabbits, then, several days later, bled the animals and mixed 0.5 c. c. of the serum with 0.5 c. c. of a solution of the respective substances which had been injected. He found in such cases that definite proteolytic action was exerted upon the injected substances by the active serum of a treated animal, whereas, in the case of most of the substances used, the normal serum possessed no proteolytic action whatever. These results were con- sistently obtained both by the dialysis and by the optical methods. It should be especially noted that the ferments studied by Abder- halden were not as specific as are the antibodies which we have dis- cussed in another place. For Abderhalden found that the serum of 8 See especially Abderhalden, Hoppe-Seyler, Zeitschr. f. physiol. Clnemie, Vols. 60, 65, and 66; also "Handbuch der biochem. Arbeitsmethoden," Vol. 5, p. 575, 1911. * Abderhalden. "Schiitzfermente," p. 49. ABDERHALDEN'S PROTECTIVE FERMENTS 495 an animal treated with proteins developed enzymes which were active, not only against the particular protein used for injection, but rather against proteins in general. They were specific only in that, when produced with proteins, they were not active against fats or carbo- hydrates. This is especially important in connection with the recent discussion concerning the identity of Abderhalden' s protective fer- ments and the specific protein antibodies. In later experiments Abderhalden showed further that similar ferments could be induced in animals by treatment with carbohy- drates and with fats. The serum of normal dogs is not capable of splitting cane sugar. However, the blood serum or plasma of a dog that has been treated with cane sugar develops the property of in- verting the cane sugar into dextrose and fructose within fifteen minutes after injection. This could easily be determined both by putting together the serum with cane sugar and determining the in- crease of reducing powers, and by means of subjecting such active plasma or serum, together with saccharose, to polariscopic examina- tion. The earlier experiments with fats were negative because the simple method of titration for fatty acids proved insufficient as an indicator of activity. However, Abderhalden succeeded in determin- ing fat-splitting properties in the blood of treated dogs by using the method of Michaelis and Rona.5 The presence of fats largely in- creases the surface tension of mixtures, and their cleavage in such mixtures consequently leads to reduction of this tension. Utilizing this principle, Abderhalden claims to have determined that the paren- teral introduction of fats into dogs is followed by a reactionary in- crease of lipases. The general significance of Abderhalden' s researches is this: When any foreign substances, protein, carbohydrate, or fats, gain entrance to the circulation of an animal, the animal body reacts by the mobilization of ferments or enzymes specifically capable of re- ducing these substances to assimilable form. It is likely that these ferments represent a mobilization of substances normally present but not concentrated in the blood stream under ordinary conditions, since they appear with a speed out of all proportion to that obtaining in the case of the antibodies discussed in another place. In one case cited by him a dog injected on November 25th, 29th, and December 4th showed powerful peptolytic serum properties on December 6th. Apparently the injection of homologous proteins into animals (i. e., rabbit serum into rabbits, etc.) does not incite reaction. These enzymes seemed to differ from specific antibodies in that they did not react solely with the substance injected, but also with other substances belonging to the same chemical group. Other dif- ferences from antibodies are the rapid appearance of the ferments 5 Michaelis and Rona. Cited from Abderhalden, loc. cit. 496 INFECTION AND RESISTANCE after treatment and their rapid disappearance after the inciting stimulus is removed. Thus Abderhalden reports that the enzymes found in a case of pregnancy disappeared within eight days after abortion or child birth. It is plain that these researches of Abderhalden offer many op- portunities for diagnostic utilization, and he has applied them to the diagnosis of pregnancy. In this condition substances from the chorionic villi get into the blood. These, according to Abderhalden, may be looked upon as in a certain sense foreign in nature, and must be chemically disintegrated by the body. In consequence it is likely that the ferments which accomplish this would appear in the sera of pregnant individuals and could be determined by his methods. When he prepared peptone from the placental substances of human beings and allowed the blood plasma of normal individuals to act upon it, observing it both by the dialysis and the optical method, no peptolytic action could be observed. However, when the plasma of pregnant women was used proteolytic action was determined. In these cases the ferment seemed to be specific for peptones produced from placental tissue both in animals and human beings, but did not act upon casein, gelatin, or other proteins. There are certain techni- cal difficulties connected with the production of a test material from the placental tissue which render this method difficult. For their more detailed description we refer the reader to the original articles. Abderhalden believes that his protective ferments may have consid- erable bearing upon the problems of bacterial immunity and anaphy- laxis, and this of course is evident to every one who has followed the development of these subjects. The problem, however, is a com- plicated one, and it is qute impossible at present to draw definite conclusions. THE MEIOSTAGMIN REACTION Ascoli and Izar 6 have attempted to work out a diagnostic reac- tion which depends upon an alteration of surface tension of a fluid when an antigen unites with its specific antibody. Ascoli in his first experiments worked with typhoid bacillus extracts and the sera of typhoid patients, and found that when the two suspensions were mixed a reduction of surface tension resulted after time for union between the two had been allowed. They determined the reduction of surface tension by Traube's 7 method by the use of apparatus spoken of as the "stalagmometer." The principle of this method depends upon the fact that as surface tension is reduced the number of drops to a given quantity of fluid is increased. 6 Ascoli and Izar. Munch, med. Woch., Nos. 2, 7, 18, 22, 41, 1910. 7 Traube. Pfluger>s Archiv, Vol. 123, 419. THE MEIOSTAGMIN REACTION 497 Diluted serum of patients was mixed with diluted antigen, and the number of drops contained in one cubic centimeter of the mix- ture was immediately determined and again measured after the mix- ture had remained for two hours in the incubator at 37° C. An example of one of Ascoli's early measurements is given in the follow- ing protocol: 1 c. c. of serum of typhoid patient diluted to 1-10. 1 c. c. alcoholic typhoid extract diluted to — ., 1 c. c. alcoholic typhoid extract diluted to — 1 c. c. alcoholic precipitate taken up in distilled „ water. . 1 c. c. in 1 0/00 alcohol in 1 c. c. 0.85 per cent. NaCl solution . . . 1 0/00 1 0/000 1 0/000 1 0/00 1 0/000 1 0/000 1 0/00 1 0/000 1 0/0000 Number of drops After Immedi- 2 hours in ately incubator 57.8 58.1 57.5 57.5 57.0 57.0 58.1 57.7 57.4 57.6 57.0 56.9 56.5 56.5 56.5 56.6 56.7 56.5 56.6 56.7 59.7 59.9 59.4 59.6 59.3 59.2 59.7 59.6 59.4 59.2 59.2 59.4 58.0 57.8 57.5 57.4 57.4 57.5 57.5 57.6 Of course a certain amount of reduction of surface tension results when various antigens are brought together with normal sera, but this can be easily controlled by suitable dilution, and must be care- fully taken into consideration in each individual case. Ascoli and Izar have applied this method to the diagnosis of tuberculosis, ty- phoid, and various other diseases, and have reported what seemed to them reliable results. So far experience with the meiostagmin reac- 498 INFECTION AND RESISTANCE tion has not been very extensive ; not all observers have been able to obtain results as apparently reliable as those of Ascoli and his col- laborators. It is not possible therefore to express a final opinion regarding this method of investigation; it contains, however, an in- teresting principle which with more exact methods of measurement may well become very important in serum diagnosis. CHAPTER XXI COLLOIDS BY STEWART W. YOUNG Professor of Physical Chemistry, Stanford University, Cal. INTKODUCTOEY IN attempting to give in the brief space of a single chapter any adequate account of the present state of our knowledge in so vast a field as that of colloid chemistry and physics one is confronted with a rather difficult problem. In the present outline the attempt will be made to get at some notion of the matter by a presentation first of the more important generalizations which have been drawn, this to be followed in each case by sufficient experimental evidence to serve as illustration, together in some cases perhaps with certain evidence which may seem to contradict in some degree such cur- rent conceptions. The reason for this particular method of pres- entation lies in the fact that new material is so rapidly accumu- lating, much of which seems more or less at variance with present accepted theories that it seems more than possible that some of these fundamental generalizations may soon undergo ma- terial modification if not reform. It would, therefore, seem ill- advised, in presenting a brief resume to readers who are not physicists or chemists, merely to present the present theories as they are used to-day by workers in the field, and to sound a note of warning that many, if not all of them, are not so securely supported by broad evidence as to allow of very concrete prediction being based upon them. Definition. — The fundamental distinction between the crystalloid and colloid states of matter was first drawn by Thomas Graham * as a result of his investigations into the phenomenon of dialysis. He noted that in general those substances which when in solution did not pass through the dialyzing membrane, or did so only very slowly, also were characterized by the fact that when they separated from solu- tion, either by precipitation or by evaporation, they did so in the non- 1 Graham. Phil Trans., 1861, 183. 499 500 INFECTION AND RESISTANCE crystalline or amorphous form. This class of bodies he named col- loids, since glue (Greek KoAAa meaning glue) presented a typical case. Colloid substances may appear in highly dispersed states, such as. dilute glue, arsenic sulphid suspensions, oil or rosin emulsions, milk (casein in highly dispersed condition), and the like, in which case they are spoken of as sols. They may also occur in the undispersed or only slightly dispersed state, as the amorphous precipitated sul- phids of the heavy metals, precipitated casein, or dry glue. In this state they are spoken of as gels. When a colloid substance has once been converted from the sol or dispersed state into the gel or undispersed state, its properties may differ greatly in different cases. Thus if a dispersed soap (soap- solution, or more correctly soap sol) be coagulated by the addition of common salt, the coagulum or soap gel may be removed from the salt solution, and if again placed in pure water it will redisperse and again assume the sol condition. Such a colloid is spoken of as a reversible colloid. If, however, an arsenious sulphid suspension be put through precisely the same course of treatment it will, in the last stage of the treatment, refuse to redisperse, and is therefore spoken of as an irreversible colloid. Some authorities prefer to speak of these two classes of colloids as emulsion and suspension col- loids, respectively,2 since in general those colloids which are reversi- ble tend to separate out in soft masses, and in general to gelatinize rather than to flocculate, while the irreversible colloids rather tend to truly flocculate and form very compact and frequently more or less granular coagula. Since, however, we seem more likely at the present time to suffer more from an excess of classification than from a lack of it, the attempt will be made to get along in this discussion with the earlier nomenclature. It may, indeed, be added that it is highly probable that the distinction between reversible and irre- versible colloids is only one of degree. For example, many of the metallic sulphids which are typically irreversible may be made to some extent reversible by means of thorough washing and re- treatment with hydrogen sulphid which had been originally used in their preparation. It is probable that certain colloids are apparently irreversible only because we do not truly reverse the conditions. Heretofore in this discussion the term "colloid'7 substance has been used as if to imply that certain chemical individuals were characteristically colloid, while others were not. It was much in this sense that Graham used the term. Investigations since his time have shown this to be a misconception, and it is now apparent that any and all substances may be either colloid or crystalloid, the form they assume depending upon treatment. Thus albumin may be crystal- lized and common salt may be obtained in the state of a colloid solu- 2 V. Weimarn. Ztschr. Chem. Ind. KolL, 1908, 3, 26. COLLOIDS 501 tion or sol.3 Albumin, gelatin, and agar may be obtained crystalline by proper regulation of temperature and the use of proper solvents, as solutions of ammonium sulphate for albumin, and alcohol-water mixtures of varying strengths for the two latter substances. Sodium chlorid has been obtained in the colloidal condition by precipitating it in a solution of sodium sulphocyanate by hydrochloric acid, each of the reacting substances being dissolved in a mixture of amyl alco- hol and ethyl acetate. There is much evidence that leads to the belief that all colloid systems are unstable. Van Bemmelen characterized them as systems which never reached a state of rest, that is, were never in equili- brium. The conditions which determine the appearance of a body in the colloid or crystalline form lead to the suspicion that bodies always separate from solution in the amorphous or colloidal condi- tion and that all crystallization is a secondary phenomenon. The conditions that are favorable for the transformation of a colloid into a crystalline form are a considerable solubility and a considerable rate of crystallization. Where either or both of these is at a minimum the conditions are favorable for relative permanence in the colloid condition. It is upon the basis of this prin- ciple that von Weimarn succeeded in obtaining relatively stable colloidal solutions of common salt and many other easily crystal- lizable salts. Furthermore Doelter 4 has succeeded in converting many well-known amorphous precipitates into crystalline bodies by means of stirring, pressure, impact, and high temperature. Among the substances thus transformed are aluminium, chromium, and iron hydroxids, and the sulphids of arsenic, antimony, and zinc. With this much by way of introduction, we may now proceed to a closer consideration of some of the better recognized properties of colloid sols and gels. For convenience we shall first take up the discussion of these systems from the more definitely physical point of view, and later take up those properties which seem more defi- nitely chemical. Physical Properties of Colloids.— 1. FORM AND SIZE. — Current opinion seems to be leading rapidly to the general acceptance of the hypothesis that in liquid systems of two or more components we have to do with a continuous series of conditions ranging from coarse sus- pensions through suspensions of increasing fineness (increasing de- grees of dispersion) to finally the molecular and ionic states of solu- tion. The opinion is also growing that, although for certain practical purposes the classification of all such systems in one way or another, as in terms of the various degrees of dispersion, may be useful, the 3V. Weimarn. Ibid., 1910, 7, 92, and "Grundziige der dispersoid Chemie," 107-108. 4 Ztschr. Chem. Ind. Koll., 1910, 7, 86. 502 INFECTION AND RESISTANCE excessive use of such classifications is likely to narrow rather than broaden our conception of the whole subject matter of the field. It would seem that the most stimulating point of view is to be reached from the acceptance of the suggestion of Wolfgang Ostwald, that the chief problem of colloid chemistry at the present time lies in deter- mining the influences of the degree of dispersion upon the physical and chemical properties of all liquid solutions, mixtures, suspensions, or what-not. If this point of view be taken it follows that the form and size of the particles in a disperse system are a matter of the first importance. If the degree of dispersion in a given system be not too great the form of the particles may be readily observed under the microscope. Such evidence shows that the spherical form predominates enor- mously over all others, although under carefully controlled conditions ovoid forms may appear, as in the case of gelatin and agar. These ovoid forms are taken by von Weimarn (loc. cit.) and others, as evi- dence of directive forces, and hence of incipient crystallization. If the system be treated in such a way as to decrease the dispersion, as, for example, if a reagent be added which tends to flocculate the col- loid, but not in sufficient quantity to produce actual precipitation, the decrease in dispersion may take place in two quite different ways : first, the size of the particles may increase, as in the case of oil emulsions; second, the particles do not coalesce but become at- tached together in chains and groups which, in many cases, resemble bunches of grapes. This sort of aggregation may go so far as to pro- duce web-like structures. The jellying of gelatin has been shown to be due to the development of such web structures. Glue shows much less tendency in this direction, and if some acetic acid be added, as in the preparation of commercial liquid glues, this web formation is almost entirely absent, and the adhesive qualities are at the same time greatly improved. It seems quite certain that both of the above modes of aggregation are possible in one and the same system at different stages in its condensation. Thus highly dispersed copper sulphid becomes aggregated in its first stages of condensation by an actual increase in the size of the spherical particles. After these reach a certain fairly definite size further aggregation takes place by the grouping together of these spheres. It is generally recognized that all grouped and webbed structures are secondary. A large number of very important investigations have been di- rected toward the determination of the size of disperse particles throughout the greatest variation in dimensions. The fact first noted by Graham, that substances in colloidal solution show a very small, and frequently almost negligible, rate of dialysis, points di- rectly to the supposition that the particles in such solutions are in a far less dispersed state than in solutions of crystalloid substances. The rate of dialysis is directly determined by the rate of diffusion. COLLOIDS 503 which, in turn, is inversely proportional to the square roots of the masses of the diffusing bodies. Measurements of osmotic pressure in solutions also give an accu- rate measure of the relative masses of dispersed systems where such measurements can be successfully carried out, and a great deal of work has been devoted to attempts to measure the osmotic pressure of colloidal solutions. Great difficulties both of experimentation and of interpretation are encountered in this field. As will soon be seen a colloid particle stands in a very complex relationship to its sur- rounding liquid, and furthermore it is a matter of extreme difficulty to obtain a colloid solution free from electrolytes, which themselves may create osmotic pressure or otherwise affect the measurements. About the only conclusion which it is safe to draw at the present time is that if colloid solutions show osmotic pressure at all the value of it is very small compared to that shown by crystalloidal solutions of substances of more or less like formula weights. This leads to the conclusion that the particles in a colloid solution are in a state of dispersion far less than that found in a typical crystalloidal solution. For a most excellent resume of the present state of our knowledge in this field the reader is referred to a recent book by Dr. L. Casuto, of Pisa, entitled "Der Kolloide Zustand der Materie" (Steinkopf, 1913). When the size of the disperse particles is sufficiently great they may, of course, be measured under the microscope, and with the advent of the ultramicroscope the limits of visibility of small bodies has been very notably extended. The ultramicroscope is known in several forms, the first having been devised by Siedentopf and Zsig- mondy. All depend upon the production of powerful rays of light in directions parallel to the surface of the microscope slide. In such a field there will be no luminosity, provided the field is optically empty, that is, contains no particles of sufficient size to produce a dispersion of light. If, on the other hand, such particles are present, the effect observed will be an illumination whose character will de- pend upon the size of the particles. If the particles are of sufficient size the illumination will show them individually as bright points of dispersion, even though the particles are too small to be observed of themselves, just as the stars are visible from the light which they disperse, but cannot of themselves be seen. If the particles are so small that they are no longer able to disperse sufficient light to make each particle appear as a bright point, there will, nevertheless, pro- vided the particles are present in sufficient numbers, be produced a diffuse luminosity throughout the field. These phenomena are wholly analogous to those observed when a beam of light is passed through a dark room in the atmosphere of which fine dust particles are found. The path of the whole beam is made apparently uni- formly luminous by the smaller particles, while occasionally there 504 INFECTION AND RESISTANCE appear points of bright illumination, due to the presence of larger particles. This is known as the Tyndall effect. The light which has passed through such fields is found to have become polarized. It is evident that, in a solution whose particles are sufficiently large to become individually visible as points of light under the ultra- microscope, it immediately becomes possible to determine the size of the particles on the assumption that these are all of the same size. The procedure consists in determining the following quantities: (1) the total number of particles in a given volume by the usual blood- count method; (2) the weight of the dispersed substance in a given volume by a chemical or other analysis; (3) the density of the dis- persed substance, which is usually taken as equal to that in the undis- persed state. This undoubtedly introduces an error in the computa- tion, since, in all probability, the density increases in the dispersed state, owing to increased compression by surface tension. This error is probably small unless the degree of dispersion is very great. By this method, particles in colloidal gold solutions have been observed and counted whose diameters were as small as 10~6 mm. This rep- resents about the limits of individual visibility under the ultramicro- scope, that is, with particles much smaller than this the field appears diffusely illuminated. This value is about one-hundredth that of the wave-length of violet light, and about ten times that of the calculated diameter of the ethyl alcohol molecule. The rate of settlement under the influence of gravity has also been used to determine the size of colloid particles. By means of Stokes' law for the fall-rate of bodies through a viscous medium, a comparatively simple equation permits of the calculation of the diam- eter of the falling body when the fall-rate, the viscosity of the me- dium, and the densities, respectively, of the dispersed substance and of the medium are known. Perrin 5 used the same principle in the preparation of suspensions in which the particles were all of more or less the same size, using, however, regulated centrifugation in- stead of simple settling under gravity. There has also been developed, largely by Bechhod 6 another method which throws some light on the relative sizes of particles, and also offers a very interesting and valuable experimental weapon for colloid investigation. This is the method of ultrafiltration. It has been found possible to produce graded filters which allow of the passage of particles below a certain size, and which restrain any larger ones. These filters are made by impregnating ordinary filters with gelatin and other colloidal solutions and drying with special precautions. The permeability decreases with the concentration of the gelatin or other substance used. 5 C. E., 146, 967, 1908. 6 Ztschr. Chem. Ind. KolL, 2, 3, 1907; Die Kolloide in Biologic u. Mcdizin, Dresden, 1912. COLLOIDS 505 The investigations as to the size of particles all lead to two gen- eral conclusions : first, suspensions and colloidal solutions in general differ from one another mainly in degree of dispersion, at least up to the limit of individual detectibility of the particles under the ultramicroscope, beyond which point at present all is speculation, although the presumption is strong and the belief is growing that there is also no other fundamental distinction to be drawn between colloidal and so-called true solutions ; second, it is always found that unless special purification is resorted to a colloidal solution contains particles of widely differing sizes side by side. 2. THE BROWNIAN MOVEMENT. — About a century ago the Eng- lish botanist, Brown, noticed that very small spores and other bodies when suspended in water, and observed under the microscope, were in a state of rapid oscillatory and rotary motion. This motion of small masses of matter has come to be known as the Brownian move- ment. It is noticed in colloidal solutions whose particles are not too large, and at the same time are large enough to be individually de- tectible under the ultramicroscope. As a result of the theoretical considerations of Einstein,7 of Smoluchowski 8 and of Corbino,9 and of the experimental researches of Svedberg 10 and of Perrin,11 the Brownian movement has come to be considered as nothing more nor less than a manifestation of that kinetic energy with which all matter is endowed, and which forms the basis of the kinetic theory of gases. A rapidly gyrating and oscillating colloid particle is there- fore looked upon as a large scale picture of the state of the molecules themselves. These investigations have probably done more than any- thing save the development of the kinetic theory itself to place molecular and atomic speculations on a firm basis of plausibility. Svedberg's investigations were instituted to determine the mean velocity of colloid particles whose mass could be determined by the ultramicroscopic method above referred to. Computing from these factors the average kinetic energy of the particles, this, according to Svedberg, gives the same value which would be computed for the particle on the basis of the kinetic theory. Perrin attacked the prob- lem from a somewhat different point of view. The number of gas molecules in the atmosphere decreases from the surface of the earth outward at a rate which is determinable by computations based on the kinetic theory. Perrin set himself the task of determining the rate of decrease in the concentration of the particles of a colloid solu- tion, in which the particles were of uniform size, the concentrations being determined at different levels in a cylinder in which the solu- 7 Ann. der Phys., 9, 417; 11, 170; 17, 549; 19, 371. 8 Ann. der Phys., 21, 756. 9 Nuovo Cimento, 20, 5. 10 Ztschr. f. Elektrochem., 12, 853, 1906. 11 C. R., 146, 967, 1908. 506 INFECTION AND RESISTANCE tion had been allowed to stand until it had reached equilibrium with the gravitational forces. The result was that the same law of dis- tribution was found to hold in this case as in the case of the atmos- phere. The kinetic theory is thus shown to apply quantitatively as wrell as qualitatively to colloidal solutions. 3. ELECTRICAL PROPERTIES. — If a U-tube be filled with water, electrodes placed in each arm, and these electrodes maintained at a constant difference of potential either by a battery, dynamo, or other source of direct current, it is noticed that there is a continual flow of liquid in the tube, in one direction near the walls and in the opposite direction in the interior of the tube. There is every reason to believe that such currents will be set up in all cases whatever the nature of the liquid or of the tube, although the current set up may in par- ticular cases be very small and even very rarely approach or equal zero. If we name the current along the walls simply the "current," and that through the interior the "countercurrent," then in the case of glass and water the direction of the current is from anode to cathode, and that of the countercurrent is from ca^iode to anode. This phenomenon is explained by the hypothesis that at the surface of contact between the glass and the water there is established a dif- ference in electrical potential, the glass becoming negatively and the water positively electrified. If this assumption is valid it follows that if a particle of glass placed in water be subjected to the influence of two electrodes placed in the water, it will, being negatively charged, be attracted by the positively charged electrode (the anode) and repelled by the negatively charged electrode (the cathode). The result would be a wandering of the particle of glass through the solu- tion toward the cathode. This result is confirmed by ample experi- ments. Furthermore, the phenomenon is common to all particles in all liquids, so far as is known, so that any colloidal solution placed in a potential gradient will show wandering of its particles in one direction or the other. Thus in water, ferric hydroxid, chromium hydroxid (and most hydroxids in the colloidal state), methyl violet, and some other dyes wander to the cathode. All colloidal metal solu- tions, sulphur, the halogen salts of silver, chlorophyll, rosin, mastic, most dyes, and, in fact, the great majority of substances investigated wander toward the anode. Albumin (and probably some other sub- stances) wanders toward the cathode in acid solution, and toward the anode in alkaline solution. As will be seen later, the hypothesis of the existence of such electrical charges on colloid particles has been of very great use in explaining many forms of conduct on the part of dispersed systems. 4. SURFACE TENSION. — If a globule of mercury be divided into two parts, these two parts will unite again if opportunity be given. All the opportunity which is necessary, if the surfaces be clean, is to bring the two parts into mechanical contact. The union of the sep- COLLOIDS 507 arate parts may, however, occur in a variety of other ways, in fact, in any way whatever whereby such union is physically possible. Thus, if the separate portions be of different size, the smaller one will have a higher vapor pressure than the larger, and evaporation from the smaller to the larger will take place until the whole of the smaller portion has transferred itself to the larger one, and the re- union is therefore complete. There is every reason to believe that if the two portions of unequal size were made electrodes in a galvanic cell, and this cell were then short circuited, that the smaller portion would go into solution and again deposit upon the larger one. In case the two portions were of the same size, these forms of recom- bination, with the exception of that of direct coalescence, would not occur if all other conditions were kept constant, but a slight differ- ence of conditions in respect to the two portions would start the act of recombination, which would then in general proceed to comple- tion. The same tendency is noticed with all substances. Thus in a liquid small crystals disappear while larger ones grow at their ex- pense, and it may be stated that, other influences for the moment ignored, the most stable configuration which can be assumed by a given mass of any substance is that in which all of the substance is in one portion, and that portion is spherical in form. This is equiv- alent to saying that all bodies so arrange themselves as to expose the least possible surface. The force which tends to bring about this condition is called surface tension. In so far as surface tension alone is concerned it follows that any colloidal solution must be un- stable, and tend to condense itself until all of the dispersed matter has aggregated itself together into a single mass of spherical form. But there are many other forces which may under certain con- ditions act against surface tension. If the dispersed substance is one that is crystalline, the directive forces of crystallization over- come those of surface tension, and the form of stable configuration will be that of the crystal instead of spherical, and equilibrium will be established when all of the available substance has aggregated itself together into one large crystal. We know, on the other hand, a great many colloidal solutions which seem to be quite stable even in very high degrees of dispersion. To explain such cases we must look for other forces working against the force of surface tension. If the dispersed particles in a colloidal solution are all charged with the same kind of electricity, they will then repel one another with a force which will vary inversely as the'squares of their distances from one another. This repulsion will then tend to work against any coalescence or other sort of union between the disperse particles. We have already seen that colloidal particles are in general charged either positively or negatively, and this may be taken to some extent as explaining the stability of such systems. Equilibrium results when the surface tension is just counterbalanced by the electrical 508 INFECTION AND RESISTANCE repulsion. The extension of this idea has been of great value in colloid investigation. The electrical repulsion will not, of course, necessarily prevent the smaller particles from dissolving and deposit- ing upon the larger ones, unless the solubility is affected by the charge. Concerning this we know nothing. The fact that colloid substances possess little or no solubility in the ordinary sense of the word means such solution and deposition must, of necessity, be a very slow process, and the colloid solution would thus appear to be perfectly stable over very long periods of time. There are many who believe that all such systems are* only apparently stable, and that on account of the absence of any sufficiently rapid means of transforma- tion which would allow the stabilizing influences to operate rapidly enough to be perceptible. Chemical Properties of Colloids — 1. It is reasonable to suppose that the chemical properties of colloid solutions are very much what is to be expected from the chemical nature of the dispersed substance as it is known under other conditions. The colloidal solutions of ar- senic sulphid should therefore react very much as would be expected of arsenic sulphid in general, except in so far as the substance is in a finely divided state in the presence of a dispersing medium (water) in which it is little soluble. Thus colloidal arsenic sulphid is soluble in alkalies and alkaline sulphid just as is the massive form. If a rod of zinc is suspended in a colloidal. solution of arsenic sulphid there takes place a slow reaction, lasting over weeks and even months, whereby the sulphur of the sulphid unites with the zinc to form colloidal zinc sulphid, while a black deposit, probably arsenic, is found on the zinc. Chemical reactions with colloids are thus, as a rule, very slow, as is to be expected, but otherwise not essentially unusual. 2. The exact chemical composition of the disperse phase in a colloidal solution is probably not definitely known in any case. In the case of colloidal metal solutions, such as gold and silver, the sus- pended particles seem to be practically pure metals, but in most cases the composition is very problematical. The 'great variation in the properties of such solutions with variations in the methods of prep- aration are undoubtedly to a great extent due to small differences in composition. Thus the properties of arsenic sulphid vary greatly with* the extent to which free hydrogen sulphid is removed from the solution, which is probably due to the differences in the amount of hydrogen sulphid absorbed or otherwise held by the arsenic sulphid. Linder and Picton believed that amorphous copper sulphid was a definite compound of copper sulphid with hydrogen sulphid. It has also been found that amorphous copper sulphid suspended in water continually deposits free sulphur, the cupric sulphid being at the same time largely converted to cuprous. It seems to be rarely or never the case that the disperse phase may be looked upon as a sub- COLLOIDS 509 stance of definite composition, being usually, if not always, a more or less complex mixture of absorption products. 3. Although not usually pure substances, it is not at all un- plausible to assume that the dispersed particles may, to some extent, undergo ordinary electrochemical ionization, in which case the par- ticles would partake of the nature of enormously large ions. This assumption is interesting as offering a purely electrochemical ex- planation of the origin of the charge which is found on such par- ticles, and it is to be said that frequently the effect of foreign sub- stances on the electrical charges of suspended particles is explain- able on this assumption. For further information on these matters reference must be had to the papers of Duclaux,12 Jordi,13 and P. P. von Weimarn.14 The Flocculation of Colloids by Electrolytes. — 1. When neutral salts are added to colloid solutions in gradually increasing amounts there always follows sooner or later a precipitation of the dispersed substance. If the salt solution be removed and pure water added in its place, this decanted, and pure water again added, so as to wash out the salt as thoroughly as possible, the final result may be that the coagulum (or gel) becomes redispersed, or such redis- persion may not occur. The outcome is determined both by the nature of the colloid and by that of the neutral salt used. 2. If acids and alkalies are the electrolytes used, the relation- ships are somewhat different. The addition of alkali to a negatively charged colloidal solution renders it more stable, while a positively charged colloid is flocculated. With acids the reverse of this condi- tion holds ; that is, the positive colloid is stabilized and the negative one is flocculated. 3. The concentration of the electrolyte required to flocculate a given colloidal solution depends very greatly on the nature of the electrolyte used. It has been generally considered that the cathion of the electrolyte is the active agent in flocculating negative colloids, while the anion is active in the case of positive ones. Thus acids (hydrogen ions) are very effective in flocculating arsenic sulphid, and alkalis (hydroxyl ions) are effective with ferric hydroxid, as might be inferred from the preceding paragraph. Different cathions, however, show very different degrees of precipitating or flocculating power. Thus in the case of arsenic sulphid, if the flocculating power of the potassium ion is taken as unity, that of calcium is about twenty, while that of the aluminium ion is three hundred and fifty. The flocculating power in this case increases very rapidly with the 12 "Thesis," Paris, 1904. 13 C. E., 136, 680, and 1,448; 137, 122; Bull. Soc. Chim., 31, 573. 14 Articles in Ztschr. Chem. Ind. KolL, 1906-1911. Also in "Grundziige der dispersoid Chemie," Dresden, 1911. 510 INFECTION AND RESISTANCE valence of the ion, a relationship which seems to be quite generally true. It has been quite commonly considered that the effect of the anion in the flocculation of negative colloids is negligible. That this point of view is not tenable has recently been strikingly shown by Sven Oden15 in his work on colloidal sulphur. He finds that the effect of electrolytes on sulphur sols, which, like arsenic sulphid, are negative colloids, is distinctly the resultant of two factors, one a floc- culating effect on the part of the cathion, the other a dispersing effect on the part of the anion. It is not improbable that our views in regard to the phenomena of electrolyte flocculation will undergo con- siderable modification in the near future, as they are based on rather scanty experimental evidence. Since the properties of sols of the same materials differ very considerably with the most minute details of their method of preparation, it is naturally difficult for the same investigator even to obtain uniform results. 4. The actual concentration of a given electrolyte required to flocculate a sol depends also very greatly on the nature of the sol itself. Some sols are precipitated by very small concentrations of electrolytes (three to four one-hundredths normal acid being usually sufficient for arsenic sulphid), while gelatin, albumin, and protein substances in general require far higher concentrations. So marked is the difference in many cases that attempts have been made to clas- sify colloids on the basis of their conduct in this respect. Thus col- loids which are very sensitive to electrolytes are called suspension colloids, while those that are not very sensitive are called emulsion colloids. To the first class belong all the true, rather coarse-grained suspensions, while the sols that yield soft gelatinous flocculates usually fall into the second class. It is also very frequently true that the latter type shows the phenomenon of reversible flocculation (see ante). This classification is for many purposes quite useful, but cannot be considered as very fundamental. For example, if the elec- trolyte used be a salt of a heavy metal most of the so-called emulsion colloids, such as albumin, are irreversibly flocculated. 5. The flocculation of sols by electrolytes is usually explained as due to the phenomenon of absorption. That is, the flocculating ion is absorbed from the solution by the dispersed particles. Since in general the ion which is absorbed is the one whose electrical charge is opposite to that of the dispersed particle the absorption results in a reduction of the charge on the particle, and allows the aggregating forces of surface tension to become operative. The evidence of the validity of this assumption is considerable. Thus the flocculated colloids always contain appreciable amounts of the ion which caused the precipitation, which is prima facie evidence of the absorption. Furthermore, the electrical charges on the particles may be meas- ured by determining their rates of migration, and the effect of elec- 15"Inaug. Diss.," Upsala, 1913, pp. 118 et seq. COLLOIDS 511 trolytes on this charge may also be observed. It is very commonly true that the addition of a precipitating electrolyte to a sol reduces materially the charge of the particles. This is not, however, always true, and the relationships are more complicated than has generally been assumed. Nor is it always true that the complete neutraliza- tion of the electrical charge on dispersed particles results in floccula- tion. For example, acetic acid may be added to arsenic sulphid sols in such quantities as to completely neutralize the negative charges of the particles and, further, so much acid may be added that the par- ticles acquire a very considerable positive charge, all this without the least signs of flocculation. Ultimately so much acid may be added as to cause flocculation. 6. In the flocculation of sols by electrolytes there is frequently observed a curious effect known as the "zone-phenomenon." It is observed when increasing amounts of certain electrolytes are added that at a certain concentration flocculation will be brought about, while if the concentration be greater flocculation will not occur, although still further increase of concentration will result in another flocculation zone. The phenomenon is most common when the electrolyte used is a salt that shows marked hydrolysis, such, for example, as ferric chlorid. If a negative sol be treated writh a solu- tion of this salt it is obvious that there will be three precipitating influences present, namely, the hydrogen ions and the colloidal ferric hydroxid, both of which are formed by the hydrolysis of the ferric chlorid and the unhydrolyzed ferric chlorid. Since the amounts of these precipitating substances in a given solution vary with the con- centration, and since each has its own concentration function in precipitation, it will be seen that the zone phenomenon may be ac- counted for in such cases. Since, however, the zone phenomenon occurs in many cases where strongly hydrolyzed electrolytes are not used, as, for example, in the agglutination of bacteria by citric and some other acids, the explanation is not wholly sufficient. There are also many curious phenomena concerned with the action of electrolytes on sols which have as yet been very little investigated, and which will probably throw considerable light on the subject. The Mutual Eeactions of Colloids — The conduct of mixtures of two different sols is of very great interest and variety, both in the absence and in the presence of electrolytes. A number of particular cases may be distinguished, and these will be taken up seriatim. 1. When two positive or two negative sols are mixed together nothing very much seems to happen. It is generally considered that the addition of one sol to another of like electrical properties results in no action. Whether this is wholly true or not is doubtful, but it is at least true that in all cases investigated up to the present time 512 INFECTION AND RESISTANCE neither individual nor mutual flocculation occurs. Whether the presence side hy side of two sorts of similarly charged disperse par- ticles results in any change in the dispersion of either is not known. Except to show that no mutual precipitation occurs, these cases have been but little studied. 2. When two oppositely electrical colloids are mixed mutual precipitation may or may not occur. The factor which in the main determines the outcome is the relationship between the amounts of the two colloids used. If neither is present in too great excess com- plete mutual precipitation in general occurs. If either is present in great excess precipitation does not in general occur. 3. The effect of a great excess of one colloid in preventing mutual precipitation is very marked when the colloid in excess is one of the emulsion-colloid type. Thus gelatin, a typical emulsion colloid, when present in excess over another colloid very frequently prevents all flocculation, even in fairly coarse-grained suspensions. Advantage has been taken of this action in preventing scaling in boilers. This scaling is due largely to the fact that lime salts held in solution as bicarbonates are decomposed by heat, with the separa- tion of calcium carbonate, at first in the highly dispersed state. This gradually aggregates together and deposits on the interior of the boiler as an amorphous flocculated colloid, which in time becomes very compact, and in many cases crystalline. If, however, a small amount of glue (impure gelatin) be added to the boiler water the colloidal constituents of the water do not flocculate and compact, but remain suspended, and may from time to time be blown off. Another illustration is found in the preparations of photographic emulsions. The silver halides flocculate very readily in pure water, but in gelatin solution remain in a highly dispersed state, which is necessary to the preparation of the plate. In this case, not only is the suspension protected from flocculation, but also a degree of dis- persion is reached which is far beyond anything attainable in pure water. The same is true when lysalbinic acid is used to prevent flocculation of colloidal silver. In pure water only very dilute sus- pensions of metallic silver are obtainable, but in the presence of lysal- binic acid suspensions containing as high as ninety per cent, of silver are obtainable, the product being used medicinally under the name of "argyrol." These are the phenomena known as "protective actions'' and the gelatin, albumin, or other colloid which exerts the protective action is spoken* of as a "protective colloid." 4. The protective action of certain colloids is not only exerted against the tendency of the protected colloid to spontaneously floc- culate, but also a certain and very great protection is offered against flocculation by electrolytes. Thus Zsigmondy 16 was able to find a definite measure of the protective action of certain colloids on the 16 Ztschr. f. analyt. Ch., 40, 697, 1902. COLLOIDS 513 precipitation of gold suspensions by means of sodium chlorid. The method was to find the amount of the protective colloid that was just necessary to prevent the flocculation of a fixed amount of a given gold sol by a fixed amount of the salt. In this way it was observed that the protective action of different colloids is very different. Thus if the amount of starch in solution which is necessary for protection be taken as 2,500 the amounts of various other colloids required are as shown in the following table : Protective colloid Starch Dextrin Gum Arabic Albumin Gelatin Glue Amount required 2,500 1,200 40 25 1 1 5. Very simple theories are devisable to explain the interactions between different colloidal solutions. Thus two oppositely electrical colloids may be considered to precipitate one another mutually be- cause of the electrical attraction existing between all oppositely charged particles. This results in bringing together the oppositely charged particles with the formation of relatively neutral aggregates, which, as shown in the discussion of the flocculation by electrolytes, is a condition favorable to precipitation. For very obvious reasons, then, no flocculation would be expected when two like charged col- loids are mixed. Many objections may be made to the unqualified acceptance of this explanation. It offers, nevertheless, a valuable leading idea when not accepted too dogmatically. A number of factors probably contribute to the protective action of many colloids. To some extent the effect may be purely mechani- cal, since increased viscosity imparted by the presence of the protec- tive emulsion colloid will shorten the mean free path of the floc- culable particles, and thus materially lessen the probability of im- pacts between them. Consequently flocculation will not occur as readily. Further, the ultramicroscope gives considerable evidence of the existence of another and very important factor. It seems certain, at least in many cases, that the protective colloid arranges itself in a film or coating around the flocculable particles, and in this way prevents the aggregation of the particles. These factors are not, however, sufficient to explain all cases of protective action. It must be considered that in some cases, at least, the protective colloid exerts a truly dispersive effect, such as would result from a nullifica- tion of surface tension forces. The ease with which a small amount of soap will emulsify a large amount of oil is difficult to explain on any other hypothesis. When an oil is shaken with pure water little or no emulsification results, wrhile in the presence of the soap as a protective colloid the same amount of work in shaking accomplishes enormously greater results. In fact, the oil will spontaneously emulsify by merely standing in contact with the soap solution. The equilibrium condition of oil in contact with pure water is reached 514 INFECTION AND RESISTANCE when the oil is very little, if any, dispersed, while in contact with soap solution equilibrium is reached only when the oil is highly dispersed. From the surface tension point of view this would be expressed by saying that in contact with pure water the surface tension is large and positive, while in contact with soap solution the surface tension is large but negative. It is im- possible to say to what extent these effects may be due to obscure chemical action. The Preparation of Colloidal Solutions — 1. This subject might perhaps upon logical grounds have best been treated in an opening paragraph. However, with the conclusions that have been reached in the foregoing discussion the whole matter may be dismissed very briefly. The conditions which must be fulfilled in order to obtain colloidal solutions are in the main summarized in the following paragraphs : 2. A medium must be chosen in which the given substance does not reach to any great extent the maximum, molecular degree of dis- persion— that is, in which what is usually called true solution does not occur to any great extent. While, as pointed out at the begin- ning of this discussion, there is no sharp line to be drawn between colloidal and true solutions, the substances that are distinctly recog- nizable at the present time as colloidal solutions carry particles which are in the neighborhood of one thousand times as large as average molecules. From media in which the dispersion is approximately molecular the dispersed substance usually shows a strong tendency to separate in the crystalline form, although it may frequently, if not generally, first appear in the colloidal form, then more or less rapidly becoming crystalline. The only distinction to be drawn is this: from solutions in which the dispersion is very great, approximating the molecular, separation in short time in the crystalline form is favored ; from solutions in which the dispersion does not approximate the molecular separation persistence in the amorphous state is fa- vored. 3. A colloidal sol or gel, one or the other, may generally be pro- duced from any substance in any medium in which the amount of molecular dispersion is at a minimum by means of any reaction whereby the new substance is produced from solution. Thus, for example, arsenic sulphid does not disperse in water to molecular extent in any considerable degree. Therefore, by mixing together a solution containing an arsenic compound which is soluble, and a solution of hydrogen sulphid, the resulting arsenic sulphid will appear in the colloidal state. Whether it appears in the dispersed state as a sol or in the flocculated state as a gel will depend mainly on the electrolyte content of the solutions which are used. If a solu- tion of arsenic chlorid be used the resulting solution will contain considerable free hydrochloric acid, and the tendency will be for the COLLOIDS 515 resultant arsenic sulphid to appear in the flocculated or gel form. If it is desired to prepare the substance in the sol form electrolytes must be avoided. This may be done by using a solution of arsenic trioxid instead of one of arsenic chlorid. Any reaction which is brought about under the above conditions will result in the formation of a colloidal product. The dialysis of salts which form insoluble hydroxids simply allows the normal reaction of hydrolysis to complete itself. The resulting hydroxids appear in such a way as to fulfill the above conditions, and conse- quently appear in the colloidal state. In this connection note the preparation of colloidal sodium chlorid (see ante). 4. In an appropriate medium most if not all substances, crystal- line or otherwise, may be brought directly, that is, without the inter- vention of specific chemical reactions, into the colloidal state. In some cases this may be accomplished by mechanical means. Thus oils violently shaken with water disperse to some extent and form emulsions of greater or less stability. By shaking glass, quartz, and the like with various liquids in which they are virtually insoluble the abrasion results in the formation of more or less dispersed systems, usually not very stable. Many metals may be brought into the dispersed state by causing an electric arc to pass between points held under a liquid. This electrical dispersion method has been very considerably used, but is obviously confined to substances which are conductors of electricity. On the other hand, many substances when merely brought into contact with an appropriate medium spontaneously undergo disper- sion. Thus gelatin, glue, tannin, and many other substances spon- taneously disperse in water. Even crystalline substances frequently do this. Thus soaps which have been crystallized from alcohol solu- tions when brought into water disperse in the colloidal state. Crys- tallized cuprous sulphid, the mineral known as chalcocyte, disperses in the colloidal form in solutions of hydrogen sulphid. Substances which go spontaneously into the dispersed colloidal state are usually spoken of as "lyophyllic" while those that tend to spontaneously leave the dispersed state are called "lyophobic." It is evident that a substance may be lyophyllic with respect to one me- dium and lyophobic with respect to another. Furthermore, a very small change in the nature of the medium may cause the change from a lyophyllic to a lyophobic colloid. Thus oils are lyophobic with respect to water, but lyophyllic with respect to even very dilute soap solutions. Applications to Biology. — 1. When one considers the relatively infrequent occurrence in biological systems of either crystalline sub- stances, or of substances that may be readily made to crystallize from water (the universal biological dispersing medium), it imme- diately becomes evident that the chemistry and physics of such 516 INFECTION AND RESISTANCE systems must be in the main colloidal. All biochemistry is thus in the main colloid chemistry. Aside from mineral salts, urea, uric acid, and a few other bodies, the reagents and products which are active in life processes are all to be found in the living organism in the colloidal state. While crystalline directive forces are in gen- eral absent, we have nevertheless to deal in biological phenomena with a great variety of directive forces of a wholly different charac- ter. Colloidal substances in high degrees of dispersion such as proteins and the like in the alimentary fluids are being continually converted into active living tissues, a process manifestly involving very definite directive forces, since the product (the tissue cells) is an organized one, even though its organization is not similar to that of a crystal. Thus the building of living tissue involves among other things the conversion of sols of many sorts (alimen- tary fluids, blood, etc.) into gels. In other words, the living tissue is to a certain extent to be looked upon as a colloidal gel, dif- fering, however, from laboratory gels in possessing a definitely or- ganized cell structure. While it is true that in the spontaneous gel formation with certain colloids, as, for example, gelatin, myelin, or web structures are formed purely as a result of the physical and chemical forces active, these cannot be said to bear any very strong resemblance to living cells. It is thus apparent that life processes differ very materially from those of the chemical laboratory. On the other hand, it is true that many of the component reactions which go to make up the life process may be very closely duplicated by labora- tory means, and that already the study from the colloid chemistry point of view of the reaction of many of the substances which go to make up the living organism has given interesting and important results. Furthermore, the whole field of colloid investigation has been greatly stimulated by the hearty support and encouragement which it has received from biologists. Some slight attempt will be made here to illustrate by a few examples the far-reaching possi- bility of explanation which 'colloid chemistry offers of certain phases of biological science. The actual accomplishment in the field is already so great that only a very' limited discussion can be offered here. 2. The action of electrolytes on emulsions of bacteria is wholly analogous to their action on colloidal suspensions. The bacterial emulsions are very sensitive to flocculation by mineral acids, one thou- sandth normal hydrochloric and sulphuric acids usually being suffi- cient to cause complete clumping and settling of the bacteria. Neu- tral salts, with the exception of those of silver, mercury, iron, and aluminium, do not flocculate the bacteria. If, however, the bacteria are first treated in the absence of electrolytes with an agglutinating serum small concentrations of salt solutions will bring about floccu- lation. It is also known that bacteria have the power of absorbing COLLOIDS 517 agglutinins from sera, so that it is evident that what we have here is a case of the production of a flocculable combination of bacteria and agglutinin, neither component of which is alone flocculable. Citric acid in concentrations ranging between one one-hundredth and one eight-hundredth normal produces flocculation. In either greater or less concentrations no flocculation is produced, which is an illustration of the so-called zone phenomenon. Like all other suspensions, bacteria are electrically charged, and consequently wander in the electric field. Under all ordinary condi- tions the charge which they carry is negative, from which in their general conduct it might be expected that they would conduct them- selves similarly to arsenic sulphid, which is also negatively charged. This is found to be the case. Their rate of migration is reduced by acids and evidently somewhart increased by alkalies, although very little alkali is necessary to bring about disintegration of the bacteria. It has also been found that all colloids are more or less sensitive to light in respect to their migration rates. Bacteria also show this phenomenon, as they migrate notably slower in the light than in the dark. It is quite possible that this reduction in their electrical charge may to some extent be responsible for the bactericidal action of light. 3. A very interesting application of the principles of colloidal precipitation by electrolytes has recently been made by Loeb.17 He finds that the eggs of the Fundulus, a small fish, are killed by being immersed in a pure isotonic salt solution, in spite of the fact that they normally develop in sea water. The factor which allows of their development must therefore be sought in some other constituent of the sea water which is absent in the pure salt solution. This Loeb finds in the presence of small amounts of calcium salts. Further, if a small amount of calcium chlorid be added to the pure salt solu- tion the eggs will develop in it as well as in sea water. Loeb's ex- planation is simple and very ingenious. He supposes that the sodium chlorid is toxic, provided it can diffuse into the egg. In the absence of calcium salts such diffusion is possible because the sodium chlorid is not a sufficiently powerful colloid precipitant to make the mem- brane about the egg impervious. It is, however, very well known that calcium salts (and all bivalent cathions) are far more effective colloid precipitants than sodium ions. Consequently the presence of a relatively small amount of calcium chlorid in the salt solution is sufficient to so condense the egg membrane as to make it impervious to the sodium chlorid, and thus render the latter non-toxic. 4. Another interesting application of colloidal principles is found in what is known as the Danysz phenomenon. It is found that the neutralization of the toxicity of diphtheria toxin by the 17 Am. J. Physiol, 6, 411, 1902. 518 INFECTION AND RESISTANCE antitoxin depends on the way in which the two are mixed. If a quantity of toxin just sufficient to neutralize a fixed amount of anti- toxin when it is added all at once be in another experiment added in small installments, the resulting mixture will be found to be still quite strongly toxic. This is quite analogous to what is found in the interaction of many colloids. The amount of a given colloid re- quired to neutralize and precipitate another depends greatly on the way in which it is added. 5. Romer,18 Field and Teague,19 and Te.ague and Buxton 20 all carried out interesting investigations directed toward determining the migration directions (electrical charges of toxins and antitoxins). Their conclusions were that all wandered toward the cathode, and that all were therefore positively charged. If this is correct the analogy between toxin and antitoxin reactions and those of simple colloids is rather mutilated, since two positive colloids are not sup- posed to react with one another. It is, however, more than possible that the above experiments are misleading. In all cases agar dia- phragms were used. Through these there would always occur a streaming of water toward the cathode as a result of the electrical potential between the agar and the water. This might well be so great as to obscure, and even more than overcome any anodic wan- dering that might occur. Furthermore, the conduct of proteins in general in the electric field is a very complex one, and one that is only just beginning to be understood. For these reasons it is at present very dangerous to draw any very dogmatic conclusions. 6. In closing mention may be made of what seems to be an im- munity phenomenon which seems rather clearly to be a case of pro- tective colloid action. It is observed when an agglutinin is added to a bacterial emulsion that if an excess of the agglutinin be added no agglutination occurs. This is wholly analogous to the fact that, while a small amount of gelatin will precipitate arsenic sulphid sus- pension, a larger amount will not. Conclusions — While great progress has been made in the field of colloid investigation from the chemical and physical sides, and while also many very striking analogies are to be found from the biological side, it is nevertheless true that we are still very much in the dark in regard to a great many matters. The one great difficulty which lies in all such investigations is that it is a matter of very great diffi- culty to duplicate results. The nature of any colloid sol or gel depends so greatly upon its whole previous history, apparently down to the least detail, that great discrepancies in experimental results are found. Even the age of a sol is frequently a matter of very great importance in determining its properties. For example 18 Berl. klin. Wochenschr., Vol. 41, p. 209, 1904. 18 Journ. Exp. Med., Vol. 9, pp. 86 and 223. 20 Ibid., Vol. 9, p. 254. COLLOIDS 519 a freshly prepared gelatin solution will not precipitate arsenic sul- phid, but it will do so after it has stood for some hours. What is now greatly needed is more data on a greater variety of colloids that have heretofore been investigated and work directed toward the preparation of colloidal solutions of definite character. Until some- thing has been accomplished in these directions all biological anal- ogies and the like cannot be anything more than qualitative, and the same holds true for many of the physical and chemical conclusions which have been discussed in this chapter. INDEX OF AUTHORS Abderhalden, 95, 98, 493- 496 Abel and Ford, 96 Abramow, 41 Abt, 387, 394 Adami, 24, 134, 234, 281, 282 Addis, 171, 303 Adler, 388 Admiradzibi, 411, 432 Altinann, 184 Amoss, 436 Anderson, 362, 365, 368, 373, 374, 376-382, 389, 390. 401, 411, 426, 428, 430, 437, 463, 464 Anderson and Goldberger, 54 Anderson and Schultz, 365 Andrejew, 372, 437 Apolant, 373 Arima, 34, 477 Aronson, 473 Arrhenius, 120, 122 Arrhenius and Madsen, 120, 121, 122 Arthus, 361, 368, 404 Arthus and Breton, 361 Asakawa, 46 Aschoff, 127 Ascoli, 148, 237, 268, 493 Ascoli and Izar, 496, 497 Auer and Lewis, 364, 390 Axamit and Tsuda, 319 Bab, 210 Babes, 65, 440 Babes and Broca, 439 Babes and Lepp, 74 Bach and Chodat, 184 Baecher, 316 Baginsky, 473 Bail, 5, 11, 20, 21, 80, 228, 229, 289, 326, 443 Bail and Kleinhans, 76 Baldwin, 442 Bandelier and Roepke, 351, 356, 357 Bang, Ivar, 44, 47, 195 Bang and Forsmann, 97 Banzhaf, 430 Banzhaf and Famulener, 379 Banzhaf and Steinhardt, 379 Barber, 15, 67 Bartel, 13 Bartel and Neumann, 343 Bauer, 209, 442 Baumann, 83 Baumgarten, 84, 135, 136 Bechold, 185, 242, 243, 266, 504 Beck, 53, 439 von Behring, 65. 73, 82, 83, 84, 85, 104, 107, 359, 390, 407, 458, 459, 472 von Behring and Kitasato, 73, 75. 86 von Behring and Kita- shima, 407 von Behring and Wer- nicke, 66, 73, 75, 86 Belfanti, 5 Belfanti and Carbone, 91 Beniasch, 246 Bergel, 204 Berghaus, 447 Bertin, 426 Bertrand, 75, 86, 105, 464 Besredka, 46, 67, 71, 75, 132, 133. 198, 349, 374, 375, 378-380, 387- 390, 401, 431, 432, 476, 484, 487 Besredka and Steinhardt, 368, 374, 377 Bessau, 402, 476 Be"sson, 289 Bickel, 323 Biedl and Kraus, 365, 368, 369, 383, 390, 402, 404 Biggs, 324, 449-451 Billitzer, 266 Billroth, 135 Biltz, 123, 242 Bispham, 319, 333 Bizzozero, 234 Blackstein, 8 Blair, 444 Blumenthal, 131 Boas, 200, 210, 211 Boehme, 139 Bogomolez, 371 Bohra, 44 Bolton, 92 Borden, J. H., 223 Bordet, 89, 91, 94, 122, 123, 140-145, 153, 154, 156, 158-160, 164, 165, 170, 186, 223, 239-243, 245, 248, 288, 296-298, 311, 416, 473 Bordet and Gay, 166, 167 Bordet and Gengou, 186, 188 Bordet and Streng, 167 Borrell, 280, 478 Bouchard, 20, 85 Boycott and Douglas, 239 Boyle, Robert, 1 Bradley, 341 Brand, 179. 180 Brau and Denier, 34, 87 Braun, 200, 205, 411 Breton, 361 Brieger, 29, 30, 77, 337, 338, 404 Brieger and Fraenkel, 73 Brieger and Mayer, 70 Briscoe, 278, 279 British Plague Committee, 479 Broca, 439 Brodie and Dixon, 369 Bronfenbrenner and Nogu- chi, 181, 183 521 Brown and Fraser, 43 Brown- Se"quard, 405 Browning, 155, 167, 177 Browning and Cruikshank, 201 Browning and McKenzie, 196 Bruck, 192, 198. 199. 205, 439. 440 Bruschettini, 41 Buchner, 33, 80, 104, 125, 134, 137, 143, 288, 300 Buchner and Hahn, 72 Buchner and Orthenberger, 178 Bujwid, 430 Buller, 286 Bullock, 341 Bullock and Western, 318 Burgers, 39 Biitschli, 294 Buxton, 518 Cagniard-Latour, 1 Calmette, 75, 86, 87, 105, 106, 174, 356. 440, 464-466, 478 Calvary, 369 Canfora, 5 Cantacuzene, 299 Carbone, 91 Carey, 183, 278. 279, 284, 306, 307, 343 Carriere, 39 Carroll, 55 Castellani, 100, 232 Casuto, L.. 503 Cattani, 84 Chamberland and Roux, 66, 73 Chantemesse, 438, 469, 475, 476 Chapin, 318, 319, 320, 323 Charrin and Roger, 89, 218 Chauveau, 57. 66, 83 Cherry, 104. 105 Chirolanza, 8 Chodat, 184 Choksy, 479 Christian and Rosenblatt, 442 Citron, 21, 101, 198, 210. 440, 469 Claypole, 8, 281 Clowes. 214 Coca, 175, 398. 405, 406, 465 Cohn, 77, 118, 145, 176, 182 Cohnheim, Otto, 258 Cole, 228, 468, 475 Cole, Docbez and Gilles- pie. 221 Cole and Meakins, 341 Collins, 457 Conradi and Drigalski, 219 Conte, 13 Contejean, 99 Corbino, 505 522 INDEX OF AUTHORS Cornet and Kossel, 60 Courmont and Doyen, 131 Cowie and Chapin, 318, 319, 320, 323 Cox, 324, 353 Craig and Nichols, 203 Craw, 124 Crile, 398 Cruikshank, 201 Cumming, 367 Currie, 428 Curschmann, 60 Dale, 398 Danysz, 18, 123, 124 Dautwitz, 97 Dean, 190, 193, 316-318, 321, 323. 480 Delanoe, 411 Delezenne, 92 Denier, 34, 87 Denys, 80, 90, 312, 326, 351, 357, 473 Denys and Havet, 168, 300 Denys and Kaisin, 300, 308 Denys and Leclef, 311, 312 Denys and Marchand, 325 Denys and Van der Velde, 73, 87 Descatello and Sturlii, 237 Detre, 199 Deutsch, 100, 301 Deutsch and Feistmantel, 326 Dick, 303 Dickson, 44, 276 DieudonnS, 447, 487 Dineur, 225 Ditman, 341 Dixon, 369 Dochez, 221 Dochez and Gillespie, 475 Doelter, 501 Doerr, 34, 87, 263, 372, 380, 410. 422, 432, 436, 469 Doerr and Russ, 381, 382, 388, 389, 391, 394, 396, 400 Doflein, 6 Dold, 413, 415, 417, 418, 421, 423 Donath, 147 Donath and Landsteiner, 169 Donitz, 46, 130 Doring, 196 Douglas, 239, 313-316, 334-337, 340 Doyen, 131 Draper and Handford, 54 Dreyer and Madsen, 113 Dreyfus, 375, 428 Drigalski, 219 Duclaux, 509 Dunbar, 434, 435 von Dungern, 92; 124, 143, 145, 171. 195, 214, 215, 263, 268, 269, 429 von Dungern and Coca, 175, 465 von Dungern and Hirsch- feld, 58, 239, 372, 436 Durham, 89, 218-220, 229- 231, 248 Dwyer, 22, 228, 310, 402, 476 Eggers, 318 Ehrlich, 37, 56, 57, 58, 75, 83, 85-87, 93, 95, 106- 120, 122, 124-126, 128- 133, 141-144, 146, 147, 149, 151, 152, 155, 156, 158-160, 162, 164, 173, 177, 182, 186, 187, 189, 234, 235, 239, 264. 301, 416, 439, 440, 443 Ehrlich and Bordet, 164 Ehrlich and Brieger, 337, 338 . Ehrlich and Marshall, 157 Ehrlich and Morgenroth, 142, 143, 146. 147, 150, 151-155, 157, 165, 177, 181 Ehrlich and Sachs, 153, 155, 165, 166 Einstein, 505 Eisenberg, 19, 229, 233, 268, 429 Eisenberg and Volk, 226, 233, 235, 236 Eisenbrey, 368, 369, 373, 39.8 von Eisler, 47, 132, 133, 195, 196 Elschnig, 437 Emery, 341 Emmerich, 85 Engelmann, 287 Epstein, 58 Epstein and Ottenberg, 239 Falloise, 171, 303 Famulener, 379 Fassin, Louise, 172 Faust, 30 Ferran, 66, 345, 351, 484 Ferrata, 178. 183 Ficker, 220, 223 Field, 518 Fildes, 210 Finsterer, 445 Fischer, 7. 128, 135 Fish, 94 Fleischmann, 99 Fleming, 324, 340 Flexner, 359, 370 Flexner and Jobling, 470 Flexner and Noguchi, 174, 465 Flexner and Sweet, 45 Flugge, 84, 87, 134 Foa, 85 von Fodor. 79, 134 Ford, 96, 234 Fornet and Miiller. 257, 261, 262, 267, 268 Forsmann, 97 Fraenkel, 24, 47, 73, 184 Frank, 92 Fraser, 43 Freeman, 340 Friedberger, 23, 38, 172, 174, 190-192, 194, 240-246, 270, 291, 366, 378, 385. 391. 396, 397, 399-402, 411, 413, 414, 419-423, 441,442 Friedberger and Hartoch, 394, 395 Friedberger and Ito, 422 Friedberger and Mita, 366, 367, 415, 430, 432 Friedberger and Moreschi, 69 Friedberger and Nathan, 418 Friedberger and Salecker, 423 Friedberger and Szmanow- ski, 418 Friedemann, 148, 242-244, 250, 265, 380-386, 390, 395, 397, 399-401, 406- 408, 442 Friedemann and Isaak, 403 Friedemann and Sachs, 176 Frosch, 5 Futaki, 19, 80, 325, 326 Gabritchewsky, 291, 311 Gaffky, 487 Galleotti, 70 Galtier, 13 Garbat and Meyer, 477 Garre, S3 Garrey, 292 Gautier, 29 Gay, 54, 68, 163, 166, 167, 189-193, 212, 438, 484 Gay and Adler, 388 Gay and Claypole, 8, 291 Gay and Rusk, 268, 269 Gay and Southard, 364, 371, 380, 382, 387- 390, 394 Gengou, 171, 172, 186, 188, 190, 192, 211, 265, 303, 304, 400, 493 German Plague Commis- sion, 13, 479 Gibier, 51 Gibson, 430, 457 Gibson and Collins, 457 Gilbert, 7, 348 Gildersleeve, 92 Gillespie, 221, 475 Glynn, 353 Glynn and Cox, 324 Goldberger, 54 Gonzales, 430 Goodall, 428 Gottlieb, 30, 39, 40, 44, 45, 99, 409 Grafe and Graham, 148 Graham, 148, 499, 502 Gramenitski, 184, 185 Grassberger, 458 Grassberger and Schatten- froh, 34, 36, 73, 87 Griffiths, 29 Grohman. 79, 134 Gruber, 80, 224, 306, 312, 326 Gruber and Durham, 89, 218-220, 229, 248 Gruber and Futaki, 19, 325, 326 Gruber and Wiener, 88 Griinbaum, 220 Guggenheim, 180 Gumaleia, 52 Gurd, 303 Guttstadt, 439 Haccius, 488 Hada and Rosenthal, 148 Haendel, 183, 194, 221, 263, 351, 373, 395, , 406, 407, 474, 475 Haffkine, 485. 486 Hahn, 55, 72, 168, 300, 304, 448, 460, 462, 480 Hall, 463 Hallier, 77 Hamberger and Moro, 391 Hammarsten, 28, 404 INDEX OF AUTHORS 523 Handford, 54 Hankin, 168, 301 Hardy, 242, 301 Harriehausen and Wirth, 462 Harris, 96 Harrison, 225 Hartoch, 394, 395 Havet, 168, 300 Hayem, 272 Hecker, 175, 180 Heidenheim, 369 Heilner, 493 Hektoen, 238. 280, 317, 319, 321, 322, 323, 325, 333 Hektoen and Ruediger, 178, 314, 318, 395 Helme, 276 Henneguy, 275 Herincourt, 74 Herman, 171, 303, 478 Herz, 477 Hewlett, 171, 303 Hirschfeld, 58, 239, 372, 436 Hiss, 218, 222,- 230, 309, 310 Hiss and Zinsser, 309 Hober, R., 44 Hodenpyl, 69 Hogyes, 66 Holobut, 411, 432 Hopkins, 239, 353 Hopkins and Ziminermann, 202 Hort, 344 Hiine, 316, 317, 321 Hunt, Reid, 409 Hunter, John, 62, 79, 134 Inmann, 316, 323 Isaak, 403 Isaeff, 85, 87-89, 137 Isaeff and Ivanoff, 218 Ito, 422 Ivanoff, 218 Izar, 415. 496, 497 Jacobams, 210 Jacobsthal, 204 Jacoby, 96, 250 Jacoby and Schiitze, 185 Jaffe, 246 Jagic, 265 Jameson, Eloise, 242 Jenner, 62, 63, 345, 481 Jennings, 286 Joachim, 227 Jobling, 332, 470, 471 Jobling and Peterson, 424 Jochmann, 306, 307, 469, 470 Jochmann and Miiller, 87 Joest, 73 Johannesen, 426 Johnston, 8, 190 Joos, 69, 226, 240, 259 Jordi, 509 Jorgensen and Madsen, 338 Kaisin, 300, 308 Kaliski, 148, 149 Kanthack, 289 Kanthack and Hardy, 301 Kantorowicz, 307 Karasawa and Schick, 448 Karlinski, 73 Karsner, 373 Kempner, 34, 73, 87 Kempner and Pollack, 41, 131 Kempner and Shepilewsky, 131 Keysser and Wassermann, 422-424 King, 375 Kiss. 176 Kisskalt, 20 Kitasato, 32, 57, 73, 75, 77, 84-86, 104 Kitashima, 407 Klausner, 204 Klebs, 272 Klein, 239 Kleinhans, 76 Klien, 332 Knaffl-Lenz 176 Knorr, 46, 125, 407 Kobert and Stillmarck, 106 Koch, 72, 76, 77, 296, 345, 355, 356, 439, 440 Koehlisch, 342 Kohn, 443 Kolle, 13, 56, 69, 83, 351, 482, 486 Kolle and Martini, 479 Kolle and Otto, 487 Kolle and Schurmann, 39 Kolle and Strong,' 487 Kolle and Wassermann, 469, 470 Korschun and Morgenroth, 169, 306 Kossel, 60, 94 Kraus. 60, 70, 87, 90, 248- 251, 365, 368, 369, 383, 390, 402, 404, 415, 421, 422. 469, 488 Kraus and Admiradzibi, 411, 432 Kraus and Doerr, 34, 87, 380, 410, 422, 432, 469 Kraus, Doerr and Sohma, 263, 372. 436 Kraus and Joachim, 227 Kraus and von Pirquet, 264-266 Kraus and Stenitzer, 477 Kraus and Volk, 200, 389 Kretz, 390. 408 Krumwiede, 52 Kruse, 20, 39 Kumagai, 402 Kyes, 174. 175, 465 Kyes and Sachs, 174 Lagrifoul, 477 Lamar, 319, 332, 333 Lambert, Adrian. 310 Lambert, R.. 277 Lambotte, 171, 303 Landois. 91, 405 Landsteiner. 47, 92, 97, 122, 169, 195, 200, 238. 250, 265, 371 Landsteiner and Dautwitz, 97 Landsteiner and Donath, 147 Landsteiner and von Eis- ler. 47, 132, 133, 195, 196 Landsteiner and Jagic, 265 Landsteiner and Levaditi, 54 Landsteiner. Muller and Potzl, 200 Landsteiner and Richter, 237 Landsteiner and Stankovic, 196, 265 Langhans, 275 Lawson and Stewart, 342 Leber, 33, 276, 248, 306 Leclainche, 36, 433 Leclainche and Vall6e, 297, 433 Leclef, 311, 312 Leishmann, 313, 315, 329, 482 Lemaire, 382 Lemoine, 267 Lenhartz, 473 LePlay, 391 Lepp, 74 Lesne" and Dreyfus, 375, 428 Levaditi, 54, 200, 322, 426, 488 Levaditi and Inmann, 316, 323 Levaditi and Yamanouchi, 204 Levin, I., 100 Lewin, 461 Lewis, 364, 374, 380, 382, 390 Lidforss, 286 von Liebermann, 175 Liefmann, 173, 174, 183 Liefmann and Conn, 145, 176, 182 Liesenberg and Zopf, 20 von Lingelsheim, 136, 178, 472 Linossier and Lemoine, 267 Lister, 79, 134 Loeb, 292, 517 Loeb, Strickler and Tuttle, 406, 407 Lo'hlein, 314 Longcope, 303, 304 Low, 287 Lowenstein, 441, 443 Lowi and Meyer, 408, 409 Lubarsch, 80, 81 Ludke, 477 Lura, 402 Lustig, 480 Lustig and Galleotti, 70 Macdonald, 319 Macfadyen, 71, 222, 477 MacKonky, 480 Madsen, 39, 107, 113. 116, 117, 120, 121, 125, 130, 337, 338, 408 Magendie, 359, 405 Magnus, 255 Malory and Wri?ht, 353 Malvoz, 224, 225 Manwaring, 167, 205, 370, 404 Maragliano, 148 Marbe", 173 Marchand, 297, 312, 325 Marfan and LePlay, 391 Marie and Levaditi, 200 Marie and Morax, 41 Marie and Tiffeneau, 133 Marinesco, 41 Markl, 479 Markl and Rowland, 469 Marks, 183, 460 Marmorek, 469, 472, 473 Marshall, 157 Martin, 190, 193, 316, 321, 453 Martin and Cherry, 105, 106 Martini, 479 Marx, 69. 100, 301 Massart and Bordet, 288 524 INDEX OF AUTHORS Mathes, 477 Matschinsky, 276 Matuso, 149 Mayer, 70 McClintock and King, 375 Mclntosh and Fildes, 200, 210 McKenzie, 196 McNeil, Archibald, 208, 216 Meakin and Wheeler, 342 Meakins, 341 Meier, 201. 204 Meltzer, 434 Mendel, 96 Mennes. 312, 325 Menzer, 473 Mesnil, 170 Metchnikoff, 31, 46, 78-81, 84, 87, 89-91, 130, 131, 136, 138, 140, 169, 170, 172, 218, 272-275, 276, 277, 296-305, 308, 484 Metchnikoff and Besredka, 67. 68, 349 Meyer, 42, 256, 320, 357, 468, 409, 440, 447, 477 Meyer and Gottlieb, 30, 39, 40, 44, 45, 99, 409 Meyer and Michaelis, 473 Meyer and Overton, 44 Meyer and Ransom, 41, 131, 447 Michaelis, 180, 246, 251, 293 473 Michaelis and Rona, 495 Michaelis and Schick, 462 Michaelis and Skwirsky, 179, 181 Miller, 92 Mironoff, 472 Mita, 366, 367, 415, 430, 432 Morax, 41 Moreschi, 69, 189, 190 Morgenroth, 86, 87, 106, 132, 142, 143, 146, 147, 150-154, 157, 165, 306, 359, 465 Morgenroth and Ehrlich, 173, 177 Morgenroth and Sachs, 363, 165, 324 Moro, 356, 391, 438, 441 Moss, 148 Mosser, 473 Mouton, 274 Moxter, 305 Much, 285, 342 Muir, 145. 190. 195 Muir and Browning, 177 Muir and Martin, 190, 192, 316, 321 Mtiller, 44. 47. 55. 87. 173, 200, 228, 250, 257, 261, 262, 267, 268, 307, 324 Mttller, Friedrich, 307, 494 Mtiller and Jochmann, 306 Myers, 251 Nathan, 418 Naunyn, 405 Neisser, 155, 156, 198,199 Neisser and Dorring, 196 Neisser and Friedemann, 242, 243, 244, 265 Neisser and Sachs, 189, 191, 212, 213 Neisser and Wechsberg, 92, 160-163, 165, 186, 191, 245 Nencki, 39, 77, 83 Nernst, 122 Neufeld, 17, 71, 75, 80, 90, 169, 290, 306, 317, 333, 344 Neufeld and Bickel, 323 Neufeld and Cole, 468 Neufeld and Dold. 413, 415, 417, 418, 421, 423 Neufeld and Haendel, 183, 194, 221, 351, 474, 475 Neufeld and Hiine, 316, 317, 321 Neufeld and Rimpau, 76, 315, 320, 321 Neufeld and Topfer, 321, 322 Neumann, 343 Nichols, 203, 234 Nicol\e, 193, 224, 250, 380, 394 Nicolle and Abt, 387, 394 Nikati and Rietsch, 54 Nissen, 80 Noguchi, 47, 164, 174, 175, 181, 183, 195, 196. 200, 203, 208, 210, 306, 438. 465 Nolf, 173. 178. 395 Norris, 251, 252 Northrup, 446 Novy, 29, 30. 463 Nuttall, 51, 79-81, 84, 87, 134, 137, 254, 255, 328 Obermeyer and Pick. 249- 251, 258, 261 Odaira, 402 Oden, Sven, 510 Ohlmacher, 432 Olmstead, 216, 217 Opie, 306, 308, 341 Oppenheim, 18 Oppenheimer, 29, 128, 250, 493 Orthenberger, 178 Osborne, Mendel and Har- ris, 96 Oschida, 490 Ostenberg, 261 Ostwald, 293. 502 Ottenberg, 208, 238, 239, Ottenberg and Epstein, 58 Ottenberg and Kaliski, 149 Ottenberg, Kaliski and Friedemann. 148 Ottenberg and Thalheimer, 148 Otto, 361, 374, 376, 380- 382, 386, 387, 487 Overton, 44. 47 Paltauf, 90, 224, 226, 232 Panum, 272 Park, 230, 349, 462 Park and Biggs, 324, 449- 451 Park and Krumwiede, 52 Park and Throne, 457 Park and Williams, 452, 453, 455 Pasteur, 1, 16, 56, 63, 64, 65, 83, 128. 136, 345, 481, 489. 490 Pasteur and Thuillier, 16 Paul, 54 Paul, Kraus and Levaditl, 488 Pauli, 241, 246 Pearce, 93. 291, 350 Pearce and Eisenbrey, 368, 369, 398 Pearce, Karsner and Eisenbrey, 373 Pearson, 344 Perrin, 504, 505 Peterson, 424 Petterson, 297, 303, 305, 308, 328 Pfaundler, 223 Pfeffer, 135, 285, 286 Pfeiffer, 33, 38, 69, 87-89, 137-140, 191, 232, 289, 324, 366, 367, 373, 378, 392. 412, 421, 444, 445, 487 Pfeiffer and Beck, 53 Pfeiffer and Bessau, 476 Pfeiffer and Finsterer, 445 Pfeiffer and Friedberger, 190, 191 Pfeiffer and Isaeff, 88, 137 Pfeiffer and Kolle, 482 Pfeiffer and Marx, 69, 100, 301 Pfeiffer and Mita, 366 Pfeiffer and Wassermann, 85, 87, 88 Philip, 118 Physalix, 464 Physalix and Bertrand, 75, 86, 105 Physalix and Contejean, 100 Pick, 29, 36, 70, 224, 249, 250, 251, 258, 261, 264 Pick and Schwartz, 250, 371 Pick and Yamanouchi, 371, 388 von Pirquet, 264, 265, 266, 356, 390, 397, 440-444 von Pirquet and Moro, 438 von Pirquet and Schick, 27, 361, 426, 427, 428 Plaut, 199 Plotz, 55 Pollack. 41, 131 Pollender, 1 Ponfick, 405 Forges, 19, 200, 236, 243, 265, 266 Forges and Meier, 200, 201, 204 Portier, 360 Portier and Richet, 360 Potter, 341, 344 Potter, Ditman and Brad- ley, 341 Potzl, 200 Powell, 353 Preiz, 19 Pribram, 87 Proescher, 220 Prudden, 80 Prudden and Hodenpyl, 69 Quincke, 294 Rabe, 97 Rabinowitch, 205 Rankin, 427 Ransom, 39, 41, 447 Ranzi, 214, 367, 373, 445 Reagh, 225, 231 INDEX OF AUTHORS 525 Rees, 353 Rehns, 225 Reid, 314. 341 Reim, 246 Rhumbler, 294 Ribbert, 234, 281 Richet, 37, 360, 380, 382, 387, 428 Richet and H6rincourt, 74, 360 Richet and Portier, 360 Richter, 237 Rietsch, 54 Rimpau, 76, 315, 320, 321 Ritz, 185 Ritz and Sachs, 182, 415 Rodet and Lagrifoul, 477 Roepke, 351, 356, 357 Roessle, 92 Roger, 89, 218, 472 Rolleston, 426, 427 Romer, 65, 101, 263, 441, 461, 462 Romer, Field and Teague, 518 Romer and Sames, 461 Romer and Somogyi, 461 Rommel and Herman, 478 Rona, 495 Rondoni, 201 Rosenau, 70, 108, 110, 351, 436, 454. 456, 457, 489 Rosenau and Amoss, 436 Rosenau and Anderson, 362, 365, 368, 372-382, 389, 390, 401, 410, 411, 426, 428, 430, 437, 463, 464 Rosenblatt, 442 Rosenow, 22. 319, 325, 326, 424. 472 Rosenthal, 148 Roser, 272 De Rossi, 225 Rouget, 289, 297 Roux, 66, 73, 125 Roux and Behring, 107 Roux and Vaillard, 104, 125, 130 Roux and Yersin, 32, 36, 72, 77 Rowland, 71, 469, 480, 487 Ruediger, 178, 314, 318, 395 Ruffer, 234 Ruppell, 356 Rusk, 268, 269 Russ, 381, 382, 388, 389, 391, 394, 396. 400 Russell, 319, 482-484 Sachs, 87, 130, 153, 155, 163, 165, 166, 174, 176, 182, 189, 190, 201, 212, 213, 324, 415 Sachs and Altmann, 184 Sachs and Kyes, 175 Sachs and Rondoni, 201 Sachs and Teruuchi, 179 Sacqu6pp6e, 228 Salecker, 423 Salmon and Smith, 20, 72 Salomonsen and Madsen, 125, 130, 337, 408 Sames, 461 Samuely, 29, 30 Sanarel'li, 85. 87 Sanpietro and Tesa, 214 Sauerbeck, 135, 136, 326 Sawtschenko, 18, 320 Schattenfroh, 34, 36, 73, 87, 304, 305 Schattenfroh and Grass- berger, 458 Scheller, 227 Schereschewsky, 203 Schick, 27, 361, 426-428, 447-449, 462 Schidorsky and Reim, 246 Schittenhelm, 370 . Schittenhelm and Weich- hardt, 403, 435 Schmidt, 204, 259-262 Schmiedeberg, 30 Schneider, 171, 303, 305 Schreiber, 459 Schucht, 199 Schultz, 365, 397, 398 Schultze, 241 Schurmann, 39 Schutze, 94, 185, 253, 255 Schwann, 1 Schwartz, 250, 371 De Schweinitz, 73, 87 Sears, 246 Sears and Jameson, 242 Selmi, 28, 77 Sewall, 464 Shattock, 237 Shepilewsky, 131 Shibayama, 19 Shiga, 219, 220 Siedentopf and Zsigmondy, 503 Siegert, 447 Simon, 332 Simm and Lamar, 333 Simm. Lamar and Bis- pham, 319, 332, 333 Simon and Thomas, 214 Skwirsky, 179, 181 Slatineau, 92 Sleeswijk, 367. 394 Smith, Alexander, 119 Smith, Henderson, 172, 449 Smith. Theobald, 9, 20, 350, 453, 454 Smith and Reagh, 225, 231 Smoluchowski 505 Sobernheim, 64 Sohma, 263, 372, 436 Southard, 364, 371, 380, 382, 386-390, 394 Sharr, 465 Stahl, 286 Stankovic, 196, 265 Steinhardt, 368, 374, 377, 379 Stenitzer, 469, 477 Stern, 80. 209 Stewart, 342 Stillmarck, 106 Stimson, 490 Strauss, 25 Strauss and Gumaleia, 52 Streng, 167 Strickler, 406, 407 Strong, 67, 486, 487 Sturlii, 237 Surmont, 92 Svedberg, 505 Sweet, 45 Swift, 201 Syme, 96 Szymanowski, 402, 418 Takaki, 46, 47, 130, 132, 133 Tamancheff, 486 Tarassewitch, 169, 301 Tarozzi, 5 Tavel, 473 Teague and Buxton, 518 Terry, 284, 461 Teruuchi, 179 Tesa, 214 Thalheimer, 148 Thomas, 57, 214 Thomsen, 444 Throne, 457 Thucydides, 61 Thuillier, 16 Tiffeneau, 133 Tizzoni, 84 Todd and White, 148 Topfer, 321, 322 Toussaint, 64, 74 Trapetznikoff, 299 Traube, 79, 134, 185, 496 Tschernorutski, 284, 343 Tschistovitch, 94, 248 Tsuda, 319 Tsuruski, 180 Tunnicliff, 325, 341 Turro and Gonzales, 430 Tuttle, 406. 407 Uhlenhuth, 70, 94, 196, 254, 255-257, 263, 267 444 Uhlenhuth and Haendel, 263, 373, 395, 406. 407 Uhlenhuth and Weidanz, 254, 268 Vaillard, 84, 104, 125, 130 Vaillard and Rouget, 289 Vaillard and Vincent, 289, 463 Vaillard, Vincent and Rou- get, 297 Valle"e, 36, 433 Van Bemmelen, 501 Van der Velde, 73, 87, 168, 473 Van Ingen, Philip, 432 Vaughan, 31, 38, 366, 385, 393, 402, 412, 413, 419, 424, 476 Vaughan, Cumming and Wright, 367 Vaughan and Novy, 29, 30 Vaughan and Wheeler, 393, 403, 419 Vedder, 177 DiVestea and Zagari, 42 Vincent, 289, 297, 463 Volk, 200, 223, 226, 233, 235, 236, 389 Wadsworth, 17, 26 de Waele, 37, 48 Wagner, 81 Waldeyer, 272 Walker, 8, 18, 172, 228, 229 303 Walker and Swift, 201 Washburn, 89, 218 Wassermann, 34, 73, 85, 87, 88, 100, 105, 106, 131, 133, 145, 155, 200, 210, 322, 349, 390, 408, 422, 423, 424, 469, 470 Wassermann and Bruck, 192, 198. 439, 440 Wassermann and Citron, 21, 101, 469 Wassermann, Meisser and Bruck, 198 526 INDEX OF AUTHORS Wassermann, N e i s s e r, Bruck and Schucht, 199 Wassermann and Plaut, 199 Wassermann and Schutze, 94. 253 Wassermann and Takaki, 46, 130, 132, 133 Webb, Williams and Bar- ber, 15, 67 Wechsberg, 92. 104, 155, 160-165, 186, 191, 245 Wechselmann, 210 Weichhardt, 392. 403, 429, 435, 436 Weichhardt and Schitten- helm, 370 Weidanz, 254, 268 Weigert, 129, 291 Weil, 76, 205, 398 Weil and Braun, 200 Weill-Halle1 and Lemaire, 382 von Weimarn, 500, 501, 502, 509 Welch, 17 Wells, Gideon, 29, 293, 294, 389, 390, 393, 403 Wendelstadt, 155 Werbitzky, 404 Werigo, 289 Wernicke, 66, 73, 83, 84, 86, 104 Western, 318 Weygant, 200 Wheeler, 393, 403, 419 White, 148 Whitfield, 341 Widal, 90, 220 Wiener, 88 Wilde, 156. 195 Williams, 15, 67, 452, 453, 455 Windsor, 313, 328 Wirth, 462 Wladimiroff, 53, 222, 252, 463 Wolff-Eisner, 38, 393, 412, 434, 440, 441 Wood, Francis Carter, 210, 221 Wright, 68, 80, 90, 313, 314, 315, 328-346, 350- 353, 367. 482 Wright and Douglas, 313, 314, 315, 316, 334- 337 Wright and Reid, 314, 341 Wright and Windsor, 313, 328 Yamanouchi, 204, 371, 373, 388, 442, 444 Yersin, 32, 36, 72, 77, 478 Yersin, Calmette and Bor- rel, 478 Yersin and Roux, 479 Young, 242, 266, 269, 287, 429, 499, et seq. Zagari 42 Zeissler, 185 Zimrnermann, 202 Zinsser, 7, 193, 305, 400, 407, 415 Zinsser and Carey, 183, 278, 279. 307 Zinsser and Dwyer, 22, 228, 402, 476 Zinsser and Johnston, 196 Zinsser and Ottenberg, 261 Zinsser and Young, 269, 429 Zopf, 20 Zsigmondy, 503, 512 Zupnik, 251 INDEX OF SUBJECTS Abderhalden, protective ferments of, 493. See also under Ferments, protec- tive, in animal body. Abscesses, secondary, caused by bac- teria, 25 Absorption theory in tox- in-antitoxin reac- tion. 123 Acne, opsonic index in vaccine treatment of, 339 Adrenal cytotoxin. 92 Agglutination, absorption experiments of Castellan! on, 232 acid, 246 value of. for differen- tial purposes, 246- 247 action of salts in, 240 agglutinability of bac- teria in, in agglu- tinoids, 229 normal differences in, between strains of same species, 228, 229 agglutinins in, 223. 224 absorption of, 232 complete, impossi- bility of, 232 heating of, 226 explanation of, 236 major, 229 normal, 233 explanation of, 234 qualitatively iden- tical with "im- mune" agglutin- ins, 234 para or minor, 229 agglutinogen in, 223, 224 effects of heating of, 226 localization of, in ec- toplasmic layers of cells. 224 agglutinoids in, 235 biological relationship and, 230 Bordet's explanation of, 240 by means of cell body proper, 225 by means of ectoplasmic substances of bac- teria, 224, 225 by means of flagella, 224. 225 cataphoresis of bacteria in, 242, 243 description of, 218 effect of gelatin addition in mastic solu- tion on. 244 V Ehrlich's interpretation of process of, 234 Agglutination, Ehrlich's in- terpretation of, diagra m m a t i c representation of, 235 Eisenberg and Yolk's in- terpretation of, 235 Ficker's reaction in, 223 flocculation of colloids , and, 241 J mechanism of, 241, 242 mutual, 242 group, 229 cause of, 230 Vdiagnostic value of, r 231 hemagglutination analo- gous to. 236, 237 history of, 218 importance of electro- lytes in, 240 in colloidal solutions, in- hibition zones in, 245 in excess of agglutinin, colloid phenom- enon and, 518 in immune serum, 141 in motile and non-mo- tile bacteria, 224, 225 in salt-free environ- ment, by addition of organic sub- stances, 245 influence of immuniza- tion with different animal species on, 232 with different species of bacteria on, 232 influence of salts on sensitized and un- sensitized bac- teria in, 243, 244 experiments of Neisser and Friedemann in, 244 inhibition zones in, 162, 236 iso-agglutinins in, 237 grouping of, 237, 238 value of presence of, 239 methods of, 218 et seq. Borden's, 223 Ficker's reaction in, 223 Gruber-Widal, 220 macroscopic, 219 microscopic, 220 Proescher's, 220 thread reaction of Pfaundler, 222, 223 nature of, 223 not associated with life of bacteria, 222, 223 527 Agglutination of "agglu- tinin" bacteria, 243 of bacteria in active im- munization, 89 of capsulated bacteria, 243 X phenomenon of, 218 power of, alterations in, by cultivation in immune serum, 228 effect of heating on, 226 spontaneous alteration in, 227 pro-agglutinoid phenom- enon in, explained as protective col- loid action, 236 pro-agglutinoid zone in, 162 pro-agglutinoids in, 235 relation of flagellar mechanism to, 222 specificity of, 219, 229 diagnostic value of, in froup reaction, 31 limitations to, 229 thread reaction of Pfaundler in, 222, 223 "two phase" theory of, 241 Agglutination reaction, diagnostic use of, 220, 221, 222 flagellar mechanism in, 222 in diagnosis of dysen- tery, 221 of glanders in horses, 222 of paratyphoid fever, 221 of pneumonia, 221 of streptococcus in- fections, 222 of typhoid fever, 220, 221 nature of, 239 et seq. precipitin reaction anal- ogous to, 263 with dead bacteria, 223 Agglutinins, 223 absorption of, in agglu- tination, 232 complete, impossibil- ity of, 233 bacteriotropins and, 321 definition of, 89 group : major, 229 para or minor, 229 heating of, effects of. 226 explanation of, 236 "immune," 234 in hemolytic serum, 93 nature of, 224 normal, 91, 233 528 INDEX OF SUBJECTS Agglutinins, normal, ex- planation of, 234 qualitatively identical with "immune" agglutinins, 234 production of, 129 Agglutinogen, 223 effects of heating on, 226 localization of, in cell body proper, 225 in ectoplasmic layer of cells, 224 in flagellar substance, 224, 225 nature of, 224 Agglutinoids, 235 Aggressins, 326 action of, 21 obtaining of, 21 secretion of, by bacteria in body, and vir- ulence, 20-22 Albuminolysins, 193, 211 Alexin, 137 a combination of soaps and proteins, 175 absence of, from aqueous humor of the eye, 170 analogy between fer- ments or enzymes and, 176 bactericidal powers of, 137 definition of, 80 dependence of, on con- centration, 176 extraction of, from leu- kocytes and lym- phatic organs, 304 et seq. filtration of. 177 in hemolysis, 144 inactivation of, by salt, 178 by salt-removal, 178 by shaking, 185 reversibility of, 184 Gramenitski's ex- periments on, 184, 185 increase of, on clot, 172 influence of salts on action of, 177, 178 leukocytic origin of, 168, 169 et seq. multiplicity of, 154 et seq. Bordet's views on, 156 Ehrlich's views on, 155 nature of, 137, 154 chemical, 174, 175 physical, 177 presence of, in blood plasma, 170-172 in blood stream, 170- 172 in normal blood, Gen- gou's view of, 303, 304 Metchnikoff's view of, 302 other experimenters on, 303 production of, in liver, 173 in thyroid gland, 172 varieties of: macrocytase, 169 microcytase, 169 Alexin fixation, 186 albuminolysin in, 193 identity of, with pre- cipitins, 193 writer's opinion on, 193, 194 Bordet and Gengou's ex- periment on, 186, 187 Bordet-Gengou phenome- non in, 188 et seq. Gay's experiments supporting, 190 in diagnosis of infec- tious diseases, 188 in diagnosis of syph- ilis, 188 Moreschi's e x p e r i - ments supporting, 189 Bordetscher Antikorper in, 194, 195 by immune animal sera with their specific antigens, 189 by protein and antipro- tein sensitizers, 189 distinguished from com- plement devia- tion, 186 during hemolysis, nature of, 176 Ehrlich's (schematic) conception of, 187 experiments of, on syph- ilitic monkeys, 198, 199 forensic tests in, 211 in anaphylaxis, 394 in determination of na- ture of unknown protein, 211 delicacy of, 212, 213 technique of, 212 in diagnosis of glanders, 216 in diagnosis of gonor- rheal infections, 216 in diagnosis of malig- nant neoplasms, 213 * von Dungern's method of, 214 antigen production for, 214 results of, 215 technique of, 214 et seq. in diagnosis of syphilis in human beings, 199 in Wassermann reaction, 198 nature of, 192 et seq. non-specific, 195 by heated normal serum, 196 by lipoidal substances in tissues, 195 by preserved normal serum, 196, 197 by protein emulsion and other ex- tracts, 196 by unsensitized bac- teria, 195 of specific precipitates, 190 et seq. albuminolysin identi- cal with, 193 Alexin fixation of specific precipitates, al- buminolysin iden- tical with. writer's opinion on, 193, 194 Gay on, 190 Pfeiffer and Fried- berger on, 190 Sachs on, 191 writer on, 193, 194 practical applications of, 198 precipitin reaction and, 190, 192 Dean on, 194 Gengou on, 192 Neufeld and Haendel on, 194 writer on, 193, 194 specific antiprotein sen- sitizers in, 192 with syphilitic serum in antigens from normal organs, 200 Alexin splitting, 178 Brand on, 179 by method of Ferrata, 178, 179 by method of Liefmann, 184 by method of Sachs and Altmann, 184 effect of acid reaction on fractions of, 181 end-piece in, 179 et seq. mid-piece in, 179 et seq. heat sensitiveness of fractions of, 180 interchange of fractions of, in different animals, 182 is it the inactivation of the mid-piece 7183 mid-piece only bound in Wassermann reac- tion, 181 physical occurrence of fractions of, in blood, 181 presensitized cell in, 180 properties of fractions of, 179 quantitative relations between fractions of, 182, 183 Alexocytes, 168 Alkali-albuminate precipi- tin, 260 "Alkalinity theory" of immunity, 83, 84 Amanita phalloides, spe- cific antitoxin from, 96 Amboceptor, 149 Bordet's definition of, 159 complementophile groups or polyceptors of (Ehrlich), 149,156 cytophile group of, 149, 152, 153 multiplicity of, 150, 151, 154 Ehrlich and Morgen- roth on, 150, 151 quantitative determina- tion of, in im- mune serum, 160, 161 specificity of, 150 INDEX OF SUBJECTS 529 Amboceptor and comple- ment, Ehrlich and Sachs's views on union of, 164, 165 Noguchi's measurement of quantitative relations of, 164 quantitative ratio be- tween, 163, 164 Amebse, artificial, 294 Anaphylactic antibody, re- lation of, to other antibodies, 400 Anaphylactic intoxication, peptone poisoning and, 404 Anaphylactic shock, 363 et seq. See also Anaphylaxis, clin- ical manifestation of effect of atropin and other drugs on, 379 Anaphylactin, 386 Anaphylatoxin, 22, 396, 413 action of alexin in, 422 with normal or in- activated immune serum, 422, 423 with salt solution, 424 inhibition of, by too vigorous and pro- longed reaction, 417 source of, 424 Anaphylaxis, 358 alexin fixation in, 394 analogy of immediate re- action in serum sickness to, 427 analogy of serum sick- ness to, 428 analogy of tuberculin re- action to, 442 anaphylactic poisoning, nature of, 403 et seq. from precipitates, 396 proteid split products of Vaughan and Wheeler in, 403 symptoms of, similar to peptone poi- soning, 404 Anderson and Schultz's work on, 365 anti-anaphylactic state in. See Antiana- phylaxis antigen in, intervals be- tween administra- tions of, 376 identity of sensitizing and toxic sub- stances of, 389 Doerr and Russ's work on, 389 Wells' work on, 389 nature of, 370 path of introduction of. 373 intracerebrally, 373 intravenously, 374 i n t r a-intestinally, 374 by feeding, 374 by rectum. 375 into large intes- tine, 375 subcutaneously, 374 quantity of, adminis- tered, 376 et seq. Anaphylaxis, antigen in, specificity of, 371 degree of. 371 organ, 372 a u t o s e n sitiza- tion in, 373 species, 372 two separate sub- stances in, 388 Doerr and Russ's experiments on, 388, 389 Gay and Adler's ex- periments in, 388 Pick and Yamanou- chi's experiments in. 388 Arthus' work on, 361 asthma and. 434 Auer and Lewis's work on, 364 autosensitization in, 373, 437 Besredka and Stein- hardt's work on, 374 et seq. Besredka's theory of, 387 Besredka's work on, 375 et seq. Biedl and Kraus's work on, 365, 368, 369 Bogomolez's work on, 371 Calvary's work on. 369 cell participation in, 390, 397 et seq. clinical manifestations of, 363 in dogs, 368 fall in blood pres- sure in, 368 fall of temperature in, 369 increase of lymph flow in. 369 intestinal reaction in, 370 lowered coagulabil- ity of blood in, 369 in guinea pig, 363 alexin reduction in, 367 circulation symp- toms in, 366 effect of atropin, and other drugs on, 365 fall of temperature in, 366 fever in, 366, 367 lowered coagulabil- ity of blood in, 367 pulmonary emphy- sema in, 364 respiratory symp- toms in. 364, 365 susceptibility of vari- ous breeds in, 368 temporary diminu- tion of polynu- clear leukocytes in, 367 in rabbits, 368 clinical significance of, 426 dependence of, on pre- ceding inocula- tion, 360 diagnostic uses of, 444 diminution of alexin af- ter anaphylactic shock in, 394 Anaphylaxis, diminution of alexin after anaph y 1 a c t i c shock in, signifi- cance of, 395 early work on, 359 Flexner's work on, 359 Friedberger's work on, 366 Friedemann's experi- ments in, in vitro. 395 fundamental principles of, 358 Gay and Southard's ar- guments against antigen - antibody reaction theory of, 387 Gay and Southard's the- ory of, 386 Gay and Southard's work on, 364 hay fever and, 434 in serum therapy. See Serum sickness in sudden attacks of ca- tarrhal nasophar- yngitis and con- junctivitis, 435 in vaccine therapy, 432 incubation time in, 360, 376 Lesne" and Dreyfus' work on, 375, 376 Magendie on, 359 Manwaring's work on, 370 mechanism of anti-ana- phylaxis in, 401 et seq. desensitization in, 401 specificity in, 403 tolerance to anaphy- lactic poison in, 402 Nicolle's theory of, 394 organ specificity in, 436 Otto's work on, 361 % Pearce and Eisenbrey's work on, 368, 369 Pfeiffer's work on, 366 phenomena related to, 405 toxic action of nor- mal sera, 405 toxin hypersusceptibil- ity, 407 Pick and Yamanouchi's work on, 371 quantitative methods ap- plied to study of, Ranzi's work on, 367 relation of alexin to, 394 relation of antibodies of, to other anti- bodies, 400 Richet and HSricourt's work on, 360 Richet and Portier's work on, 360 Rosenau and Anderson's work on, 362, 374 et seq. sessile receptors, theory of. 390 specificity of, 362, 363 sympathetic ophthalmia and, 437 Theobald Smith phenom- enon in, 361 Theobald Smith's work on. 363 toxic action of normal sera and, 405 530 INDEX OF SUBJECTS Anaphylaxis, toxin hyper- susceptibility and, 407 transference of, 362, 379 et seq. true antigen-antibody re- action, 390 tuberculin ophthalmo-re- action and, 440 tuberculin reaction and, 438. See also Tu- berculin reaction tuberculin skin reaction and, 440 Vaughan and Wheeler's theory of mechanism of, 393 Vaughan and Wheeler's work on toxic fraction of pro- tein molecule in, 393 Vaughan's work on, 366, 367 Weichhardt and Schit- tenhelm's work on, 369, 370 with bacterial extracts, 363 with normal serum, 362 with proteins, 362 Anaphylaxis, bacterial, 410 anaphylatoxin formation in, 413. See also under Anaphyla- toxin. difference of speed of re- action of sensi- tized and unsen- sitized bacteria in, 418 endotoxin theory of pro- duction of, 412 Friedberger's experi- ments in, 413, 414 facts deduced from, 415 effect of excess of bacteria adminis- tered in, 415 effect of excess of sensitization on anaphylatoxin in, 415 effect of too pro- longed exposure in anaphylatoxin in, 416 nature of bacterial in- fections and, 419 et seq. Neufeld and Dold's ex- periments in, 417 serum anaphylaxis and, 411 Vaughan's theory of bac- terial splitting as cause of, 412, 413 Anaphylaxis, passive, 379 et seq. Biedl and Kraus's work on, 383 Doerr and Russ's work on, 380 duration of, 381 Friedemann's work on, 380, 381, 383 Gay and Southard's work on, 380 interval between injec- tion of sensitized serum and injec- tion of antigen in, 382 Anaphylaxis, passive, methods of pro- duction of, 380, 381 nature of reaction of, 382 Nicolle's work on, 380 Otto's work on, 380 Richet's work on, 380 Weill-Halle" and Le- maire's work on, 382, 383 Anderson and Schultz's work on anaphy- laxis, 365 Anthrax, relative suscepti- bility of man and animals to, 53 study of, in regard to resistance and immunity, 296 vaccination against, his- tory of, 64 method of, 64 Anthrax bacilli, attenua- tion of virulence of, 18 virulence of, 15 Anti-alexin, 157 action of, 157 Anti-amboceptor, 152, 153 Anti-anaphylaxis, 362, 377 Besredka's work on, 375 et seq. mechanism of, 401 et seq. desensitization in, 401 tolerance to anaphy- lactic poison in, 402 methods of producing, 377 Besredka and Stein- hardt's methods, 377 Rosenau and Ander- son's methods, 378 specificity of, 378, 403 "Anti-antibodies," 147 Antibodies, concentration of. in lymphatic organs in active immunity, 100 in other organs in ac- tive immuniza- tion. 101 . in active immunization, 85 locality of production 'of, dependent on locality of anti- gen concentra- tion, 101 normal, explanation of, 234 origin of, 100 specificity of, 85 Antibody formation, body cell in, 125 chemical nature of, 126 chemical action of anti- gens in, 128 in active immunity, re- moval of spleen and, 100 mechanism of (side chain theory), 124 by internal secretion of body cell, 125 processes of metabol- ism and, 125 overproduction of recep- tors in, 128 Antibody formation, prin- ciples of, 94 Anticomplement, 157 action of. 157 Anticytophile interpreta- tion of anti-sensi- tization (Ehrlich and Morgenroth), 153 controversy on, 153 "Antiformin," 70 Antigen-antibody reactions, 129. See also Toxin - antitoxin reaction. antibody production in body cells in, 130 relationship between susceptibility of tissue and toxin- binding properties in, 131 side chain theory in, 129 specificity of, 97, 129 variety of antibodies in, 129 agglutinins, 129 antitoxins, 129 cytotoxins, 129 precipitins, 129 Antigenic properties of cells, relation of, to lipoid constit- uents, 97 Antigens, action of, 95 active immunization with, analogies to drug tolerance, 99 characteristics of, vari- ations in, 98 complex structure of, Ehrlich and Mor- genroth's concep- tion of, 151 definition of, 35, 94 "local" immunity in or- gans directly in contact with, 101 locality of production of antibodies depen- dent on locality of concentration of, 101 organ specificity of, 98 protein nature of, 96 specificity of, 97 Anti-isolysins, 147 "Antiricin," 85 Antisensibilisin, 387 Antisensitization, anti- cytophile interpre- tation of (Ehrlich and Morgenroth), 153 controversy on, 153 Antisensitizers, 152, 153 non-specificity of, 154 Antitoxic serum, direct ef- fect of, on toxin, 104 indirect protective ac- tion of, against toxin, 104 "normal" serum of Behr- ing, 107 Antitoxin, chemical rela- tions of, with toxin, 114 definition of, 85, 86 diphtheria. See Diph- theria antitoxin. INDEX OF SUBJECTS 531 Antitoxin, production of, 129 by true toxins, 35 snake venom, 464 effect of heat on, 105 specific, substances in- citing, 86 standardization of, 463 by means of toxin, 107 guinea pigs used in, 108 tetanus, production of, 463 use of, in passive im- munization, 86 Antitoxin unit, diphtheria, 107 Antivenin, 464 Arrhenius and Madsen on neutralization in toxin - antitoxin reaction, 120 Arthus, phenomenon of, 380 work of, on anaphylaxis, 361 Ascoli and Izar's work on meistagmin reac- tion, 496 Asiatic cholera, relative susceptibility of man and animals to, 53 Asthma, anaphylaxis and, 434 Atrepsie, 56 Attenuation of bacteria by chemicals, 66 by cultivation under pressure, 66 by drying, 65 by heating, 65 by passage through ani- mals, 65 by prolonged cultivation above optimum temperature, 65 by prolonged growth on artificial media in presence of own metabolic prod- ucts, 65 capsule formation in, 18 Auer and Lewis's work on anaphylaxis, 364 Autocy totoxins, 93 . - 6 "J Autogenous vaccines, 351 "A u t o hemolysins," 146, 147 Auto-inoculation by mas- sage in active immunization, 340 Autolysins, 146 Autosensitization in ana- phylaxis, 373 Auxilysin, 167 Avian tuberculosis, relative susceptibility of animals to, 52 Bacillus botulinus, action of, 4 Bacteria, adaptation of, in body, 6, 7 agglutinability of, alter- ations in. 226 by cultivation in immune serum, 228 Bacteria, agglutinability of, caused by heating, 226 normal differences in, between strains of same species, 228, 229 spontaneous, 227 in agglutinoid, 229 agglutination in. See under Agglutina- tion. "agglutinin" bacteria, agglutination of, 243 aggressin secretion of, in body, and vir- ulence of, 20-22 anti-opsonic properties of, and antichem- otactic sub- stances, 325 artificial cultivation of, 10 attentuation of, 18 methods of. 65. See also under Atten- uation. by laboratory ma- nipulations, 17 capsulated, agglutination of, 243 virulence of, 326 capsule formation in, and virulence, 18 colloid phenomena and action in, 516 destruction of, by cy- tases in leuko- cytes, 301, 302 by exudates, 300 by phagocytes, 300 different strains of, va- riation in infec- tion from, 15, 16 ectoplasmic hypertrophy of, in relation to virulence, 19. 20 effect of body tempera- ture on invasive powers of, 12 effect of cultural adap- tation of, on vir- ulence, 12 effect of path of intro- duction of. on infection, 12-14 on virulence of, 12-14 effect of quantity of, in- troduced, on in- fection, 14 entrance of, into body tissues, 6 generalized action of, 24 growth of, within leu- kocytes, 298 in blood stream, 24 in localized infection, re- action to, 26 through accidental conditions. 25 in normal serum, resis- tance to phagocy- tosis of, 325 incubation of. 26 localized action of, 23 measurement of relative degrees of viru- lence of, 15 negative charge of, in suspension, 242 number of, introduced, and relative viru- lence, 15 Bacteria, occurrence of, 2 parasitic and saprophy- tic, classification of. 11 phagocytosis of. See under Phagocyto- sis. relative virulence of dif- ferent strains of same, 15, 16 resistance of, to phago- cytosis, due to nonabsorption of opsonin, 326 resistance of living cell to, 6 secondary abscesses caused by, 25 selective action of, in localized infec- tion. 25 selective lodgment of, in tissues, 40 sensitized, immunization with, 68 sensitized and unsensi- tized, influence of salts on aggluti- nation of, 243, 244 similar conditions pro- duced by differ- ent, 23 specificity of, and infec- tion, 22 use of, in active immu- nization, 85, 87- 89 variation in virulence of, when successively passed through animals. 16, 17 Bacterial anaphylaxis. See Anaphylaxis, bac- terial. Bacterial extracts, active immunization with, 69 extraction of bacteria for, by mechan- ical methods, 71 by permitting them to remain in fluid media, 70 Bacterial infections, con- ceived as reac- tion of body against a foreign antigen, 420 nature of (Friedberger), 419 et seq. Bacterial precipitins, group reactions in. 251 diagnostic value of, 252 partial or minor, 252 Bacterial products, active i m muni zation with, 72 Bactericidal properties of blood serum, 134 Bacterial proteins, 33 Bacterial toxins. See also Toxins. action of, after distribu- t i o n in body, 40 active immunization with, 72 chemical structure of, in relation to tox- icity, 43 endotoxins. See Endo- toxins. 532 INDEX OF SUBJECTS Bacterial toxins, lesions produced by, in course of excre- tion, 45 local injury by, 45 nature of, 32 nature of union of, to body cells, 44 nerves attacked by, 41 obtaining of. 32 from dead cultures, 32 from living cultures, 32 physical relationship with body cells in action of, 44 production of antitoxin by, 35 selective action of, 40 principles of, 43 reasons for, 45 selective localization of, 39 specific distribution af- ter introduction of, 40 specific susceptibility of tissues to, 45 chemicals inhibiting, 47 true, 33 analogy of, with en- zymes, 36 bacteria producing, 34 characteristics of, 34 heat sensitiveness of, 36 incubation time of, 36 Bactericidal powers of blood serum, 79 alexin in, 137 nature of. 137 in vivo, 137 cholera experiments in, 137 Bactericidal substances, origin of, from leukocytes, 169 et seq. Bacteriolysins, agglutinins and, 321 in active immunization, 89 Bacteriolysis, 137 extracellular theory of, 140 heat a factor in, 138 mechanism of, 138 immunity conferred by, 137, 138 in immune serum, 137, 138 Bordet's findings in, 140, 141 inactivation and re- activation i n , 140 Pfeiffer's phenomenon in. technique of, 138 et seq. specificity of p r o - tection of, 137, 138 transference of power of, 137, 138 leukocyte action in, 168 leukocytes in, 140 Bacteriotropins, bacteri- cidal sensitizers and, 321 et seq. normal opsonins and, 320 et seq. Bacteriotropins, presence of, in immune sera without ly- sins, 322 specificity of. 321 thermostability of, 320 Bail's aggressin theory, 21, 67 classification of para- sites, 11 Bauer's modification of Wassermann test, 209 Baumgarten's osmotic the- ory of bactericidal powers of blood, 135 Behring, Kitasato and Wernicke, a n t i- body theory in ac- tive immunity of, 84 Besredka and Steinhardt's work on anaphy- 1 a x i s , 374 et seq. on serum therapy of ty- phoid fever, 476 Besredka's anti-endotoxic serum in treat- ment of typhoid fever, 476 method of administra- tion of antitoxin, 431 theory of anaphylaxis, 387 vaccines in prophylactic immunization against plague, 487 work on antianaphylax- is, 375 et seq. Biedl and Kraus's work on anaphylaxis, 365, 368, 369 on passive anaphylaxis, 383 Blood, non-putrefaction of, 79, 134, 256 phagocytic activities of, in immunity, 79 Blood plasma, cell-free, inhibition of bac- terial growth in, in immunity, 79 presence of alexin in, 170-172 Blood serum, agglutina- tion in, immune, 141 .klexin in, 137 nature of, 137 antibacterial powers of, in immunity, 79 anti-isolysins in, 147 autohemolysins in, 146, 147 bactericidal and agglu- tinating powers of, Wright's stud- ies of, 328 bactericidal action of immune. 81 alexin in, 137 early theories regard- ing, 134 in natural immunity, 79, 80 in vivo, 137 cholera experiments in, 137 mechanism of. 135 Blood serum, bactericidal action of immune, mechanism o f , assimilation the- ory of, 136 by chemically un- favorable environ- ment, 135 osmotic theory of, loo bacteriolysis in immune, loY Bordet's findings in, 140, 141 cholera experiments in, 137 summary of facts in, 138 heat a factor in, 138 mechanism of, 138 immunity conferred by, 137, 138 inactivation and reac- tivation in, 140 intracellular theory of, 138 leukocytes in, 140 Pfeiffer's phenomenon in. technique of, 138 et seq. specificity of protec- tion of, 137, 138 bacteriolytic powers of, transferable, 137 cell-free, bacterial growth in. 81 hemolysinogens in, 148 hemolysis in immune, 141 alexin or complement in, 144 analogy of, to bacter- iolysis, 142 Bordet's work on, 141 Ehrlich and Morgen- roth on mechan- ism of, 142 haptophore groups in, 142 relation of antigen, amboceptor and complement in, 143-145 work of Ehrlich and Morgenroth on, 143-144 work of Liefmann and Cohn on, 145 isohemolysins in. 146 protective action of, against bacteria, 50 Blood stream, presence of alexin in, 170-172 Body fluids, bactericidal powers of, in nat- ural immunity, 80 Bogomolez's work on ana- phylaxis, 371 Borden's method of agglu- tination, 223 Bordet, explanation of, on agglutination, 240 findings of, on bacterio- lytic power of immune serum, 140. 141 o n neutralization i n toxin-antitoxin re- action, 122 views of, concerning re- lation of antigen, amboceptor and complement, 158,. 159 INDEX OF SUBJECTS Bordet, views of, concern- ing relation of antigen, a m b o - ceptor and com- plement, action of complex in, 159 schematic representa- tion, 159 Bordet-Danysz phenomenon in neutralization in toxin- anti- toxin reaction, 123 Bordet and "Gengou's ex- periment on alex- infixation.186,187 Bordet-Gengou phenomenon in alexin fixation, 188 et seq. Gay's experiments sup- porting, 190 Moreschi's experiments supporting, 189 Botulinus poisoning, ac- tion of, 41 Botulinus toxin, 4 Bouillon Filtre (Denys), 357 Bovine colloid, 167 Bovine tuberculosis, rela- tive susceptibility of man and ani- mals to, 52 Brownian movement in colloids, 505 Buchner on bactericidal power of blood in natural immunity, 80 Calmette's investigations in snake poisons, 464 et seq, ophthalmo-reaction, 440 Capsule formation in bac- teria, by attenua- tion, 18 virulence and, 18 Calvary's work on anaphy- laxis, 369 Carriers, bacillus, 2 Castellani, absorption ex- periments of, in agglutination, 232 Catarrhal nasopharyngitis and conjunctivi- tis, sudden at- tacks of, anaphy- lactic nature of, 435 Cell receptors, overproduc- tion of, 152 Cellular theory of immu- nity. 136 Cerebrospinal meningitis, epidemic, serum therapy in, 469 early investigations in, 469 Flexner and Jobling's work on, 470 Jochmann's investiga- tions in, 469, 470 Kolle and Wassermann's investigations in, 469 nature of action in, 471 results of, 470. 471 standardization of serum in, 471 Chantemesse's early ex- periments in se- rum therapy of typhoid fever, 475 Chemotaxis, 285 anaphylatoxin and, 291 Engelmann's studies in, 287 influence of bacteria in, 288 influence of bacterial extracts in, 288, 289 malic acid in, 286 of slime-molds or myxo- mycetes, 286 of spermatozoa of ferns, 286 Pfeffer's technique in, 286 physical explanations of, 292 et seq. selective, 291, 295 surface tension in, 293 "artificial amebae" and, 294 Chicken cholera, vaccina- tion against, his- tory of, 63 Cholera, active prophylac- tic immunization against, 484 Ferran's early investi- gations in, 484, 485 Haffkine's method in, 485 Kolle's method in, 486 Strong's method in, 486 Asiatic, relative suscep- tibility of man and animals to, 53 chicken, vaccination against, history of, 63 effect of path of intro- duction of bac- teria of, on infec- tion, 14 experiments in. showing bactericidal pow- ers of blood se- i rum in, 137 hog, immunization with bacterial products in, 72 Cobra antitoxin, action of, 465 standardization of, 466 Cobra lecithid, 175 Cobra venom, action of, 465 Coctoprecipitin. 258 "Coefficient of extinction," 332 Cole's work on sen m therapy of pne i- monia, 475 Colloids, 499 application of phenom- ena of, in elec- trical field, 518 to action in animal body, 516 to action in bacteria, 516, 517 to action with eggs of Fundulus, 517 to biology, 515 et seq. to Danysz toxin-anti- toxin phenome- non, 517 to nonagglutination in excess of agglu- tinin, 518 chemical properties of, 508, 509 Colloids, chemical proper- ties of, chemical composition in, 508 chemical reactions in, 508 electrochemical ioniza- tion in, 509 Classification of, 500 definition of. 499 emulsion, 500, 510 flocculation of, by elec- trolytes, 509 acids and alkalies in, 509 concentration of elec- trolyte in, 509 explanation of, 510 nature of sol in. 51O precipitin reaction an- alogous to. 265 salts in, 509 zone phenomenon in, 511 gel, 500 Graham's work on, 499 irreversible, 500 lyophobic, 515 lyophyllic, 515 mutual reactions of, 511 in two oppositely elec- trical sols, 512 explanation of, 513 in two similarly elec- trical sols, 511 protective action of electrolyte in, 512 protective action of great excess of one colloid over the other, 512 explanation of, 512 nature of, 500 physical properties of, 501 et seq. Brownian movement of particles in, 505 distribution of parti- cles in, 505, 506 electrical properties in, 506 form of particles in, 502 kinetic energy in, 505 size of particles in, measurement of, 502 et seq. microscopic, 503 osmotic pressure in, 503 rate of settlement in, 504 ultrafiltration meth- od in. 504 surface tension in. 506 et seq. preparation of solutions of, 514, 515 reaction in. analogous to complement devi- ation phenomenon of Neisser-Wechs- berg, 162 inhibition zones in, 162 reversible, 500 sol, 500 stability of, 501 suspension, 500, 510 Complement. See also Alexin. amboceptor and, Nogu- chi's measurement of quantitative relations of, 164 534 INDEX OF SUBJECTS Complement, amboceptor and, quantitative ratio between. 163, 164 union of, Ehrlich and Sachs's views on, 164, 165 definition of, 144 in hemolysis, 144 multiplicity of, 154 et seq. Bordet's views on, 156 Ehrlich's views on. 155 nature of, 354 chemical, 174 Complement deviation, 160 et seq. argument in favor of Bordet's views, 162, 163 colloid reactions, analo- gous to, 162 Gay's explanation of, 163 in hemolytic reactions, 163 Morgenroth and Sachs's experiments sup- porting, 163 pro-agglutinoid zone re- action analogous to, 162 Complement fixation, 186. See also Alexin fixation. in determination of na- ture of unknown protein, 211 delicacy of, 212. 213 technique of, 212 practical applications of, 198 test of, in diagnosis of glanders, 216 in diagnosis of gonor- rheal infections, 216 in diagnosis of malig- nant neoplasms, 213 von Dungern's method of, 214 antigen produc- tion for, 214 results of, 215 technique of, 214 et seq. Complement splitting, 178. See also under Alexin, splitting of. Complementoid, 158 Complementophile group of amboceptor. 149 Conglutinin, 167 Conjunctiva, susceptibility of, to infection, 13 Corpus luteum cytotoxin, 92 "Cryptogenic tetanus," 5 Cultivation of bacteria, ar- tificial, 10 Cytases, in phagocytes, de- struction of bac- t e r i a by, 301, 302 Cytolysins, 92 Cytolytic substances, ori- gin of, from leu- kocytes, 169 et seq. Cytophile group of ambo- ceptor, 149, 152, 153 Cytotoxins, 92 specificity of, 92 Danysz effect in neutral- ization in toxin- antitoxin reac- tion, 123 Danysz toxin-antitoxin phenomenon, ap- plication of col- loid phenomena to, 517 Daphnia, Metchnikoff's study of, 274, 296 phagocytosis in, 296 Dean's antiplague sera, 480 Denys and Leclef on im- portance of phag- ocytosis in im- munity, 311. 312 Diphtheria, active immu- nization in, with toxin - antitoxin, 458 relative susceptibility of man and animals to, 53 Diphtheria antitoxic ser- um, normal, 107 Diphtheria antitoxin, 446 antitoxin production in, 455 concentration of, 457 presence of, in blood of normal individ- uals. 448 preservation of, 108 speed in administration of, 447 speed in absorption of, 449 on intravenous injec- tion, 449, 450 on subcutaneous in- jection. 449, 450 speed of diagnosis for, necessity of, 451 stability of, 108 standardization of, 455 by means of toxin, 107 early attempts in, 107 statistics showing reduc- tion of mortality with, 446 toxin production for, 452 choice of culture in, 452 cultivation of strain in, 452, 453 culture medium in, 453 "maturing" of toxin in, 453 testing of toxin in, 454 Theobald Smith's method of, 454 unit of, 107 Diphtheria bacillus, action of, 4, 5 Diphtheria toxin, action of, 40 construction of, 118 determination of diph- theria immunity with, 462 normal, 107 stability of, 108 Diphtheria toxin, unit of, 107 Diphtheria toxin-antitoxin, neutral mixtures of, 458 Behring's method of im- munization with, 458 advantages of, 459 chief value of, 459 danger of anaphylaxis in, 459 determination of pres- ence of free tox- in or antitoxin in convalescents fol- lowing treatment with, 462 human susceptibility to, 459 production of homolo- gous antitoxin in human beings with, for passive i m m u n i zation, 460 results of treatment with, 460 standardization of anti- toxin, 460, 461 limes-necrosis of toxin in, 461 Romer's method of, 461 toxic action of, 458 Doerr and Russ's experi- ments on two sep- arate substances in anaphylactic antigen, 388 Doerr and Russ's work on passive anaphy- laxis, 380 Dochez and Gillespie's work on serum therapy in pneu- monia, 475 Drug tolerance, analogy between, and ac- tive immunization with antigens, 99 Dunbar's work on hay fever, 434 von Dungern's method of alexin fixation in diagnosis of malignant tu- mors, 214 antigen production in, 214 results of, 215 technique of, 214 et seq. "Dust cells" of the lungs in phagocytosis, 279 Dysentery, agglutination reaction in diag- nosis of, 221 Ehrlich, conception of alexin fixation (schematic) o f, 187 of relation of antigen, amboceptor and complement, 149, 150 interpretation of agglu- tination by, 234 diagrammatic repre- sentation of, 235 INDEX OF SUBJECTS 535 Ehrlich, on multiplicity of alexin or comple- ment, 155 side chain theory of, in toxin - antitoxin reaction, 124 Ehrlich's "antiricin," 85 Ehrlich and Morgenroth on multiplicity o f ainboceptor, ex- ample of, 150, 151 Ehrlich-Sachs phenomenon i n sensitization, 165 Bordet and Gay's inter- pretation of, 166, 167 Eisenberg on residue an- tigen and anti- body in precipitin reaction, 268 Eisenberg and Volk's in- terpretation o f agglutination, 235 Endocarditis, malignant, 24 Endolysins, 305 Endotoxins, 33, 34 characteristics of. 37 toxic cleavage products of, 38 Engelmann's studies in chemotaxis, 287 Enzymes, analogy of, with true bacterial toxins, 36 in phagocytosis, endo- cellular and ex- tracellular, 305 Epithelioid cells, action of, in phagocytosis, 284 Epitoxoids. definition of, 112 Erythrocyte laking. 91 "Exhaustion theory" of immunity, 83 Exotoxins, 33. See also Toxins, true, bacteria producing, 34 characteristics of, 34 chemically indefinable nature of, 35 Fermentation, infectious disease and, 1 micro-organisms caus- ing, 1 Ferments, protective, in animal body, 493 Abderhalden's experi- ments with, 494 significance of, 495 diagnostic value of, in pregnancy, 496 difference of, from anti- bodies, 495 leukocyte origin of, 494 methods of determining presence of, in blood, 494 dialysis method, 494 optical method, 494 Ferran's investigations in active prophylac- tic immunization against cholera, 484, 485 Ferrata, experiments of, in complement splitting, with salt solution, 179 Ficker's reaction in agglu- tination, 223 Flexner's observations on anaphylaxis, 359 on serum therapy in cerebros p i n a 1 meningitis, 470 Forensic alexin fixation tests, 211 Forensic determination of unknown pro- teins, 211 delicacy of, 213 technique of, 212 Fornet and Miiller, ring test of, for pre- cipitin blood tests, 257 Friedberger, experiments of, in bacterial anaphylaxis, 413, 414 on the nature of bac- terial infections, 419 et seg. work of, on anaphylaxis, 366 Friedemann, experiments of, on anaphylaxis in vitro, 395 work of, on passive ana- phylaxis. 380. 381 in rabbits, 383 Fundulus, Loeb's experi- ments with eggs of, 517 Garbat and Meyer's work on serum therapy of typhoid fever, 477 Gastric juice, action of, on stomach itself, 6 Gastro-toxin, 92 Gay and Adler's experi- ments on two separate sub- stances in ana- phylactic antigen, 388 Gay and Southard's objec- tions to antigen- antibody theory of anaphylaxis, 387 theory of anaphylaxis, 386 work on anaphylaxis, 364 on passive anaphy- laxis, 380 Gay's sensitized killed vac- cines in prophy- lactic typhoid fever immuniza- tion, 484 Gels, 500 Giant cells in phagocy- tosis, 280 foreign body, 280 tuberculous, 280 Glanders, alexin fixation test in diagnosis of, 216 in horses, agglutination reaction in diag- nosis of, 222 relative susceptibility of man and animals to. 53 Gonococcus, relative sus- ceptibility of man and animals to, 53 Gonorrheal infections, alexin fixation test in diagnosis of, 216 Gottstein-Mathes' work on serum therapy of typhoid fever, 477 Graham's work on colloids, 499 Gramenitski's experiment in reversal of alexin inactiva- tion by heating, 184, 185 Grofcman on inhibition of bacterial growth by cell-free blood plasma in immu- nity, 79 Gruber-Widal reaction in diagnosis, 220 Haffkine's early work on prophylactic im- m u n ization against plague, 486 method in active pro- phylactic immuni- zation against cholera, 485 Haptines, 129 of the third order, 150 varieties of, 129 Haptophore group in toxin, action of, 110 Haptophore groups in hemolysis, 142 Hay fever, anaphylaxis and, 434 Dunbar's study of, 434 reaction in, 434 anaphylactic nature of, 435 toxic nature of, 435 treatment of, 435 Heat-alkali-precipitin, 259 Heat-precipitins, 259 Hemagglutination, agglu- tination of bac- teria by serum analogous to, 236, 237 Hemoglobinuria, paroxys- mal, 147 autohemolysins in, 147 hemolysis in, 147 Hemolysinogens, human, 148 nature of, 148 Hemolysins, anti-isolysins in, 147 autolysins in, 146, 147 isohemolysins in, 146 specific, definition of, 92 specific inciting of, in animal, 91 Hemolysis, alexin or com- plement in, 144 amboceptor in, action of, 149 et seq. anti-amboceptor in, 152, 153 anti-cytophile interpre- tation of anti- sensitization in (of Ehrlich and Morgenroth), 153 controversy on, 153 antisensitizer in, 152, 153 anti-isolysins in, 147 INDEX OF SUBJECTS Hemolysis, experiments of Liefmann and Cohn on, 145 in immune serum, 141 analogy of, to bacteri- olysis, 142 Bordet's work on, 141 Ehrlich and Morgen- roth on mechan- ism of, 142 haptophore groups in, 142 relation of antigen, amboceptor and complement in, 143-145 multiplicity of ambocep- tor in, 150, 151, 154 recapitulation of views of Ehrlich and Morgenroth on, 152 Hemolytic properties of normal serum, 91 Hemolytic reactions, com- plement deviation in, 163 Hemolytic serum, agglu- tinins in, 93 Ehrlich and Morgen- roth's conception of neutralization of, by antilysin or anti-ambocep- tor reacting with cytophile group, precipitins in, 94 Hemolytic substances, ori- gin of, from leu- kocytes, 169 et seq. Hepatotoxin, 92 Hiss, investigations of, on therapeutic use ofl leukocyte ex- tracts, 309. 310 Hog cholera, immunization with bacterial products in, 72 Hogyes method of treat- ment in rabies, 492 Holobut's work on bac- terial anaphy- laxis, 411 Hopkins' method of stand- ardization of vac- cines, 353 "Horror autotoxicus," 147 Human isolysins. See Iso- lysins, human Humoral theory of immu- nity, 136 Hydrophobia, active pro- phylactic immu- nization against, 489. See also under Rabies Hypersusceptibility. See Anaphylaxis toxin, and anaphylaxis, 407 Immune serum. See also Serum, immune, agglutination in, 141 bacteriolytic power of, transferable, 137 bacteriolytic properties of, Bordet's find- ings in, 140, 141 Immune serum, direct neutralization of toxin - antitoxin reaction, the pro- tective power of, 124 hemolysis in, 141 alexin or complement in, 144 analogy of, to bac- teriolysis, 142 Bordet's work on, 141 Ehrlich and Morgen- roth on mechan- ism of. 142 haptophore groups in, 142 relation of antigen, amboceptor and complement in, 143-145 work of Ehrlich and Morgenroth on, 143-144 work of Liefmann and Cohn on, 145 phagocytosis in, 90 specific, agglutination of bacteria in, 89 precipitin formation in, 90 Immune isolysins, 148 Immunitats Einheit, 107 Immunity, acquired, 60 artificially, 63 definition of, 62 history of, 61 increased phagocytosis and, 299 active, relation of pha- gocytosis to, 329 cellular theory of, 136 definition of, 3 diphtheria, determina- tion of, with diphtheria toxin, 462 "high tide" of, 340 humoral theory of, 136 lasting, diseases in which one attack conveys, 60 diseases in which one attack does not convey, 61 local, in organs directly in contact with antigens, 101 in skin infections, 102 natural, 49 cellular theory of. 80 definition of, 50, 62 humoral theory of, 80 inflammation in, 78 mechanism of, 78 theories concerning, 78-82 bacterial destruc- tion by phago- cytic cells, 78 bacterial growth in cell -free blood serum, 81 bactericidal pow- er of blood in natural immuni- ty, 80 bactericidal pow- er of normal blood in nat- ural immunity, 79 immunity, natural, mech- anism of, theories concerning, bac- tericidal proper- t i e s of extra- vascular plasma or serum, 81 inhibition of bac- terial growth by cell-free blood plasma, 7 9 intracellular de- struction of bacteria, 78 phagocytic activ- ities of blood, 79 phagocytic activities of blood in, 79 principles of, 50 body temperature in, 51 cultural conditions for bacteria in body a factor in, 56 increased invasive powers of bac- teria in. 56 individual differ- ences in, 58 inheritance in, 56 racial differences in, 55 relative resistance of animals in, 51 species resistance in, 51 Pfeiffer phenomenon in, 138 phagocytosis in, 90 Immunization, 60 against snake venoms, 464 history of, 61 Immunization, active. See also Vaccine ther- apy. against anthrax, 64 against chicken cholera, 63 against small-pox, 62 agglutination of bacteria in, 89 "alkalinity theory" of, 83, 84 antibacterial, 85, 87 antibodies in, 85 bodies in, fundamen- tal principles of theory of, 84 origin of, 100 antitoxic, 85 as a therapeutic meas- ure, action of, in generalized sys- temic infections, 347 in local infections, 346 in successive local infections, 347 value of. 346 in acute diseases, 350 in subacute or chronic cases, 349 as prophylactic measure, value of, 345, 346 autogenous vaccines in, 351 auto-inoculation by mas- sage in, 340 INDEX OF SUBJECTS 537 Immunization, bacteria used in, 85, 87-89 bacteriolysins in, 89 by means of living but attenuated cul- tures, 65 concentration of anti- fa o d i e s in lym- phatic organs in, 100 in other organs in, 101 definition of, 63 "exhaustion theory" of, 83 "high tide" of immunity in, 340 in diphtheria, with tox- in-antitoxin mix- tures, 458. See also Diphtheria toxin-antitoxin. invasion of bacteria in, mechanism of re- action in tissue cells against, 102 locality of production of antibodies depen- dent on locality of antigen con- centration i n , 101 negative phase in, 338 second injection in, 338 successive inoculations in, 338, 339 summation of, 338, 339 non-bacterial antitoxin- stimulating sub- stances in, 86, 87 "osmotic theory" of, 84 phagocytosis in, 90 phenomena following, 82 precipitin formation in, 90 reaction of tissue cells to invasion in, 102 removal of spleen in, and antibody-for- mation, 100 "retention theory" of, 83 second positive phase in, 339 specificity of antibodies in, 85 summation of positive phase in, 339 tuberculins in, 355 vaccines in, 351 production of, 351 sensitized. 355 with dead bacteria, 351 with living bacteria, 351 standardization o f, 353 Hopkins' method of, 353 Wright's method of, 352, 353 with antigens, analogy between drug tol- erance and. 99 with bacterial extracts, 69 extraction of bacteria for, by mechan- ical methods, 71 Immunization, with bac- t e r i a 1 extracts, extraction of bac- teria for, by per- mitting them to remain in fluid media, 70 with bacterial products, 72 with dead bacteria, 68 methods used in kill- ing bacteria for, ing 68 with fully virulent cul- tures in sublethal amounts, 66-68 with sensitized bacteria, 68 Immunization, active pro- phylactic, in man, 481 against cholera, 484. See also under Chol- era. against plague, 486. See also under Plague against rabies, 489. See also under Rabies against small-pox, 488. See also under Small-pox against typhoid fever, 482. See also under Typhoid fe- ver Immunization, passive, 74 antitoxins in, 86 definition of, 64 history of, 74 in diphtheria. See Diph- theria antitoxin in diseases caused by bacteria which do not form soluble toxins, 466. See also under Serum therapy. therapeutic application of, 75 toxin-antitoxin reaction in, 104 underlying principles of, 75 Immunized animals, bac- teriolysis in. 137 summary of facts in, 138 Incubation of bacteria, 26 Infection, acquired resis- tance to, 60 adaptation of bacteria in tissues in, 6, 7 aggressin secretion of bacteria in body and, 20-22 body temperature and, ol capsule formation of bacteria and, 18 chronic, adaptation of bacteria in, 8 clinical manifestations of. 28 conjunctiva susceptible to. 13 criteria governing, 3 cultural conditions for bacteria in body and, 56 defence of intestinal tract in, 12 defence of mucous mem- branes in, 12. 13 defence of skin in, 12 Infection, definition of, 5, 6 different, produced by same bacteria, 23 effect of body tempera- ture on invasive powers of bac- teria in. 12 effect of cultural adap- tability of bac- teria on, virulence of, 12 effect of path of intro- duction of bac- teria on, 12-14 effect of quantity of bacteria intro- duced on, 14 entrance of bacteria in body tissues in, 6 focus in, 7-9 from bacteria in blood stream, 24 generalized, 24 increased invasive pow- ers of bacteria a factor in, 56 incubation of bacteria in. 26 individual differences and, 56 inheritance and resis- tance to, 56 localized, 23 reaction in, 26 selective action of bacteria in, 25 through accidental conditions, 25 natural resistance against, 49 of various diseases, rel- ative susceptibil- ity of man and animals to, 52 protective action of blood serum against, 50 protective action of leu- kocytes against, 50 protective action of tis- sues against, 50 ptomains and, 28 ptomains as indirect cause of, 31 racial differences and, oD resistance of living cell to, 6 secondary abscesses in, secondary modifying fac- tors in, 2 selective lodgment of bacteria in body and, 40 similar, produced by dif- ferent bacteria, 23 species resistance to, 51 specificity of bacteria and, 22 susceptibility to, racial differences in, 55 relative, 51 variation in. of differ- ent strains of same bacteria, 15, 16 variation in degree of, in bacteria suc- cessively passed through animals, 16, 17 538 INDEX OF SUBJECTS Infection without infec- tious disease, 6, 7 Infectious disease, defini- tion of, 6, 8 Inflammation, process of, and phagocytosis, 280 et seq. with pyogenic staphylo- cocci, 281 with tubercle bacilli, 283 Influenza, relative suscep- tibility of man and animals to, 53 Inheritance, a factor in re- sistance to infec- tion, 56 iso-agglutinins in blood serum influenced by, 58 Inhibition zones in colloid reactions, 162 in precipitation and ag- glutination, 162 of sera in agglutination, 236 Intestinal tract, defence of, in infection, 12 Iso-agglutinins, 237 grouping of, 237, 238 in blood serum, 58 value of presence of, 239 Isohemolysins, 146 Isolysins, human, 148 grouping of, 148 testing of, for trans- fusion, 149 iso-agglutinins analogous to, 237 Isoprecipitins, 255 Jacobsthal's ultramicro- scopic method of finding precipi- tates in syphilitic sera, 204 Jenner, Edward, experi- mentation of, for i m muni zation against small-pox, 62 Jobling's work on serum therapy in cere? brospinal menin- gitis, 470 Jochmann's investigations in serum therapy of epidemic cere- brospinal menin- gitis, 469, 470 Kolle's method of prophy- lactic vaccination in cholera, 486 Kolle and Otto's investi- gations in pro- phylactic immuni- zation against plague. 487 Kolle and Wassermann's investigations in serum therapy of epidemic cerebro- spinal meningitis, 469 Kraus, Rudolf, discovery of specific precip- itins by, 248 Kraus and Doerr's study of bacterial ana- phylaxis, 410, 411 Kraus and Stenitzer's se- rum in treatment of typhoid fever, L+, definition of, 109 method of determination of. 109 Lo, definition of, 109 constancy of. 110 method of determination of. 110 Laking, erythrocyte, 91 "Landsteiner phenomenon" of autohemolysis in hemoglobin- uria, 147 Leishmann's technique for determination of opsonic index. 329 "Leistungskern," definition of, 126 Leprosy, relative suscepti- bility of man and animals to, 54 Lesne" and Dreyfus' work on anaphylaxis, 375, 376 Leukine, 305 Leukocyte extracts, thera- peutic use of, 308 et seq. Leukocytes, alexin extrac- tion from, 304 et seq. growth of bacteria in, 298 in bacteriolysis, 140 action of, 168 in leukocytosis, action of, 275, 276 in phagocytosis, 324 origin of bactericidal and hemolytic substances from, 168, 169 et seq. phagocytic powers of, 50 proteolytic enzymes from, in phagocy- tosis, 306, 307 Leucocytosis, 290 bacteria decreasing, 290 bacteria increasing, 290 sources of leukocytes in, 290 Leukoproteases, 306, 307 Leukotoxin, 92 Limes-necrosis (L-n), 461 L i p o i d constituents of cells, relation of, to antigenic prop- erties, 97 Lister on phagocytic activ- ities of blood in natural immunity, 79 Liver, production of alexin in, 173 Loeb's experiments with eggs of Fundulus, 517 Lubarsch on bactericidal properties of ex- travascular plas- ma or serum in immunity, 81 Ludke's work on serum therapy in ty- phoid fever, 477 Lustig's antiplague se- rum, 480 Lysins, production of, 130 Macrocytase, 169, 301 Magendie on anaphylaxis, 359 Malta fever, relative sus- ceptibility of man and animals to, 53 Manwaring's work on ana- phylaxis, 370 Markls serum in treat- ment of plague, 479 Marmorek's work on serum therapy of strep- tococcus infec- tions. 472, 473 Measles, relative suscepti- bility of man and animals to, 54 Meat poisoning, 4, 31 Meistagmin reaction, 496 Ascoli and Izar's experi- ments in, 496, 497 value of, in diagnosis, 497 Meningitis, epidemic cere- brospinal, serum therapy of, 469. See also under C e r e b r ospinal meningitis, epi- demic Metabolism, processes of, compared with those of antibody formation, 125 Metchnikoff and Besredka's living sensitized vaccines for pro- phylactic typhoid immunization. 484 Metchnikoff on bacterial growth in cell- free blood serum, 81 theory of, on bacterial destruction by phagocytic cells in natural immu- nity, 78 Metchnikoff 's soured milk therapy, 31 Microcytase, 169, 301 Minimum lethal dose, defi- nition of, 108, 109 method of determination . of, 109 MLD, definition of, 108, 109 method of determination of, 109 Morgenroth's toxin - HC1 modification in t o x i n-antitoxin reaction, 106 Mucous membranes, de- fence of, in infec- tion, 12, 13 Mushroom, specific anti- toxin from, 96 Narcotics, reduction of phagocytosis by, 299 Natural immunity. See Immunity, nat- ural "Negative" phase in active immunization, 338 second injection in, 338 successive inoculations in, 338, 339 "summation'' of, 338, 339 INDEX OF SUBJECTS Neisser and Friedemann, experiments of, on influence of salts on sensitized bacteria in agglu- tination, 244 Neisser and Wechsberg, phenomenon of, 160 et seq. analogous to colloid re- actions, 162 argument in favor or Bordet's views, 162 Gay's explanation of, 163 Morgenroth and Sachs experiments sup- porting, 163 pro-agglutinoid zone re- action analogous to, 162 Neoplasms, malignant, alexin fixation in diagnosis of, 213 von Dungern's method of, 214 antigen production for, 214 results of, 215 technique of, 214 et seq. Nernst on views of Arrhen- ius and Madsen on neutralization in toxin-antitoxin reaction, 122 Neufeld and Dold's experi- ments in bacterial anaphylaxis, 417 Neufeld and Haendel's work on serum therapy of pneu- monia, 474 Neurotoxin, 92 in snake venom, 465 Nicolle's theory of ana- phylaxis, 394 work on passive anaphy- laxis, 380 Noguchi's modification of the Wassermann test, 208, 209 schematic presentation of, 209 "Normal" diphtheria anti- toxic serum, 107 "Normal" diphtheria tox- in, 107 "Normal" serum. See un- der Serum agglutinins in, 91 hemolytic properties of, 91 opsonins in, 91 toxic action of, and an- aphylaxis, 405 Nuttall on bactericidal power of normal blood in natural immunity, 79 Nuttall's experiments on determining zoo- logical classifica- tions by means of precipitin reac- tion, 254, 255 Ophthalmia, sympathetic, Elschnig's expla- nation of, as ana- phylactic r e a c- tlon, 437 Opium, reduction of pha- gocytosis by, 299 Opsonic action, phagocy- tosis due to. 313 Opsonic index, determina- tion of, Leish- mann's technique for, 329 Simon, Lamar and Bispham's tech- nique of, 332, 333 Wright's technique for, 330 et seq. difficulties in, 332, 333 value of, 333 fluctuation of, in un- treated patients under influence of exercise of dis- eased parts, 340 in autoinoculations by massage, 340 in sera of normal and infected individ- ual, comparison of, 334 in serum therapy, com- parison between that in exudate of infected foci and blood serum, 340 in staphylococcus infec- tions, 334 during vaccine treat- ment with dead s t a p hylococcus cultures, 335 In treatment of gonor- r h e a 1 arthritis with autoinocula- tion by massage, 340 in vaccine therapy, im- provement and, 341 of acne, 339 of staphylococcus fu- runculosis, 335 of sycosis, 336, 337 of tuberculosis, 341- 343 value of, in control- ling therapeutic vaccinations, 344 in showing degree and conditions in which vaccination is successful, 344, 345 vaccine therapy and, 328 et seq. value of, in therapeusis, 338 Opsonic powers of normal serum, 314 reduction of, by heat, 314 Opsonins. See also Pha- gocytosis definition of, 313 immune, bactericidal sensitizers and, 321 et seq. increase of, 315 heated, increase of power of, by ad- dition of fresh normal serum, 318 reactivation of, by addition of alex- in, 318 normal and, 320 et seq. Opsonins, immune, resist- ance of, to heat, 315 specificity of, 321 thermostability of, 320 normal, 91 cooperation of heat- stable and heat- sensitive body in, 319 instability of, 314, 315 nature of. 316 similarity of, to alex- in or complement, 316, 317 specificity of, 318 production of, in thyroid gland, 173 qualitative difference be- tween normal and immune, 315, 316 specific thermostable, in normal serum, 317 Organ specificity of anti- gens, 98 "Osmotic theory" of im- munity, 84 Otto's work on anaphy- laxis, 361 in passive anaphylaxis, 380 Pancreas cytotoxin, 92 Panum's theory of intra- cellular destruc- tion of bacteria in natural immu- nity, 78 Parasites, biological transi- tion of sapro- phytes to, 5 Bail's classification of, 11 Parasitic bacteria, 4 Paratyphoid fever, agglu- tination reaction in diagnosis of, 221 Paroxysmal hemoglobin- uria, 147 hemolysis in, 147 Partial absorption method of E h r 1 i c h in measurement of toxin - antitoxin combination, 115 Pasteur, "exhaustion the- ory of," 83 experimentation of, on i m m u n i zation against chicken cholera, 63 work of, on immuniza- tion against an- thrax, 64 on prophylactive im- munization in ra- bies, 489 Pathogenic bacteria, adap- tation of, in tis- sues, 6, 7 entrance of, in body tis- sues, 6 saprophytic nature of certain, 4 Pathogenic micro-organ- i s m s, definition of, 3 occurrence of, 2 resistance of living cell to, 6 540 INDEX OF SUBJECTS Pearce and Eisenbrey's work on anaphy- laxis, 368, 369 Persensitized cells, 180 Petterson's investigations on therapeutic use of leukocyte ex- tracts, 308 Pfaundler's thread reac- tion in agglutina- tion, 222. 223 Pfeiffer on causes of bac- terial anaphylax- is, 412 work of, on anaphylaxis, 366 "Pfeiffer phenomenon" in active immuniza- tion, 89 in bacteriolysis, tech- f, nique of, 138 et seq. Metchnikoff's view of phagocytosis in peritoneal exu- date and, 302 Phagocytes 276 fixed, 276 macrophages, 277 microphages, 277 motile, 276 Phagocytosis, 272 acquired immunity and, 298 a 1 e x i n extraction in, from leukocytes and lymphatic or- gans, 304 et seq. chemotaxis in, 285 influence of bacteria in, 288, 289 influence of bacterial extracts in, 288 malic acid in, 286 of slime-molds or myx- omycetes, 286 of spermatozoa of ferns, 286 Pfeffer's technique in, 286 destruction of bacteria in, 297, 300 by alexin (or cytase) in leukocytes, 301, 302 action of, 302 Metchnikoff's inter- pretation of, 302 by exudates, 300 by phagocytes, 300 destruction of red blood cells by, 276 differences in degree of, due to bacteria, 325 differences in phagocytic energy in, due to leukocytes in, 324 differences in virulence of bacteria, de- pendent on their resistance to leu- kocytes in, 325 digestion among proto- zoa and, 274 "dust cells" in, 279 early investigations in, 272 endothelial cells in, 278, 279 enzymes in, endocellular and extracellular, 305 eosinophile cells in, 278 Phagocytosis, fixateur or sensitizer in, ac- tion of, in im- munized animals, 301 giant cells in, 280 foreign body, 280 tuberculous, 280 in daphnia, 296 in higher animals, 296 in immune serum, 315 bacteriolysins in, bac- tericidal sensitiz- ers and, 321 et seq. heated, opsonic action in, increase of, by addition of fresh normal serum, 318 increase of, 311 attributed to "stim- ulins," 311 with addition of leukocytes, 312 opsonin contents a factor in, 313 opsonins in, increase of, 315 normal opsonins and, 320 et seq. specificity of, 321 thermostability of, 320 in immunity, 90 in normal serum, op- sonins in, cooper- ation of h e a t- sensitive body in, 319 nature of, 316 similarity of, to alexin, 316, 317 specific thermosta- ble, 317 specificity of. 318 in process of inflamma- tion, 280 et seq. with pyogenic staphy- lococci. 281 with tubercle bacilli, 283 increase of, by injection of leukocyte ex- tracts, 308 et seq. in increased resist- ance, 329 with acquisition of immunity, 299 intracellular digestion and, 274 in vertebrates, 275 leukocytes in, 324 action of. 275, 276 polynuclear, 278 leukocytosis in, 290 bacteria decreasing, 290 bacteria increasing, 290 lymphocytes in, large, 278 measure of degree of, in active immuniza- tion, 329 Leishmann's technique for, 329 Simon, Lamar and Bispham's tech- nique for, 332, 333 value of, in therapeu- sis, 338 Wright's technique for, 330 et seq. Phagocytosis, measure of degree of, Wright's tech- nique for. diffi- culties in. 332, 333 value of, 333 mechanism of process of, 280 et seq. Metchnikoff's early in- vestigations o n, 273, 274 normal and immune op- sonic action in, quantitative dif- ferences between, 324 normal degenerative and retrogressive proc- esses and, 276 observation of. in vitro, 313 of micro-organisms, with or without cul- ture media, 297 opsonins in. See Op- sonins phagocytes engaged in, varieties of, 276 fixed, 276 macrophages, 277 microphages, 277 motile, 276 process of inflammation in, 280 et seq. proteolytic enzymes from leukocytes in, 306, 307 qualitative difference be- tween normal and immune opsonic substances in, 315, 316 reduction of phagocytic activity in, 298 by growth of bacteria within leukocytes, 298 by protection of bac- teria from pha- gocytes, 299 by use of narcotics, 299 relation of, to active im- munity, 329 relation of virulence to, 312 removal of extravasa- tions of blood and, 275 resistance of bacteria to, due to non- absorption of op- sonin, 326 resistance of infected subject and, 296, 297 resistance of virulent bacteria to, in normal serum, 325 spontaneous, 313 tissue cells in, 278 varieties of body cells engaged in, 278 dependent on nature of invading sub- stance, 279 Pick and Yamanouchi's ex- periments on two separate sub- stances in ana- phylactic antigen, 388 work on anaphylaxis, 371 INDEX OF SUBJECTS 541 von Pirquet and Schick's studies of serum sickness, 427 et seq. von Pirquet's tuberculin skin reaction, 440 Placentar cytotoxin, 92 Plague, active prophylactic i m m u n i zation against, 486 Besredka's vaccines in, 487 Haffkine's early work on, 486 Kolle and Otto's in- vestigations i n, 487 Rowland's vaccine in, 487 Strong's investigations in, 487 relative susceptibility of man and animals to, 53 serum therapy of, 478 Dean's serum in, 480 Lustig's serum in, 480 Markl's serum in, 479 Rowland's serum in, 480 value of, 480, 481 Yersin. Calmette and Borrel's investi- gations in, 478 Yersin's serum in, 478- 480 value of, 479 "Plasmines," 72 Pneumococcus infection, relative suscepti- bility of man and animals to, 54 Pneumococci, mutation of, 472 Pneumonia, agglutination reaction in diag- nosis of, 221 serum therapy in, Cole's work on, 475 Dochez and Gilles- pie's work on, 475 nature of action in, 474, 475 Neufeld and Haendel's work on, 474 Poison-ivy, specific anti- toxin from, 96 Pollantin, 435 Poliomyelitis, relative sus- ceptibility of man and animals to, 55 Polyceptors (Ehrlich), 156 Precipitation, 248 inhibition zones in, 162 Precipitin reaction, 248 against heated proteins, 258 et seq. coctoprecipitin in, 258 experiments on, 260- 262 h e a t-alkali-precipitin in, 259, 260 native precipitin in, 259 70° precipitin in, 259 Schmidt's experiments on, 260, 261 agglutination reaction analogous to, 263 analogy of, to colloidal flocculation, 265 autocytotoxins in, 263 Precipitin reaction, bac- terial precipitins in, partial or minor, 252 specificity of, 251, 252 Ehrlich's conception of, 264 electrolytes in, effect of, 265 forensic blood test in, 257 ring test of Fornet and Miiller in, 257 group reactions of bac- terial precipitins in, 251, 252 diagnostic value of, 252 heat precipitins in, 259 heated precipitating ser- um, effect of mixed sera in, 266-267 protective action of, 266 inhibition zones in, 265, 266 isoprecipitins in, 255 medico-legal value of, 254 non-specific partial reac- tions in, elimina- tion of, 254 Nuttall's experiments on determining zoo- logical classifica- tions by, 254, 255 organ specificity in, 262, 263 precipitinogen in, 249 chemical nature of, 249 effect of heating on, 258 non-protein, 249, 250 obtaining of. 249 precipitinoids in, for- mation of, 265 precipitins in, delicacy of, 253 determination of po- tency of. 253 inactivation of, by heat, 264 effect of, in bac- terial filtrates, 264 production of, against unformed p r o- teins, 252, 253 methods of, 251 of specific, 249 by pepton, 251 effect of heating on, 249 in animal sera by foreign protein, 248 structure of (Ehrlich). 264 zymophore group in, 264 effect of heat on, 265 quantitative proportions in, effect of, 265, 266 relative concentration of reacting bodies a factor in, 265, 266 residue antigen and an- tibody in, 267 et seq. explanations of, 268 et seq. Precipitin reaction, residue antigen and anti- body in, experi- ment on, 269 salts in, effect of, 265 species determination by means of, 253, 254 species specificity in, 262 specificity of, 248, 253, 254 vegetable proteins deter- mined by, 255 zoological classifications by means of, 254, Precipitin tests, methods of performing, 255 et seq. forensic blood test in, 257 ring test of Fornet and Mtiller in, 257 Precipitinogen, 249 chemical nature of, 249 effect of heating on, 258 non-protein, 249, 250 nature of. 250 obtaining of, 249 Precipitinoids, 265 Precipitins, 248 against heated proteins, 258 et seq. coctoprecipitin, 258 experiments on, 260- 262 heat - alkali-precipitin, 259, 260 native precipitin, 259 70° precipitin, 259 Schmidt's experiments on, 260, 261 bacterial, group reac- tions in, 251 partial or minor, 252 specificity of, 251, 252 definition of, 90 delicacy of, 253 determination of potency of, 253 heat, 259 in hemolytic serum, 94 inactivation of, by heat, 264 effect of. in bacterial filtrates, 264 isoprecipitins, 255 organ specificity of, 262, 263 production of, 129, 249 against unformed pro- teins, 252, 253 methods of, 251 specific, by pepton, effect of heating on, 249 in animal sera by foreign protein, 248 "species," specificity of, 262, 263 specific, 248 discovery of, by Ru- dolf Kraus, 248 structure of { Ehrlich), 264 zymophore group in, 264 effect of heat on, 265 Pregnancy, diagnostic value of Abder- halden's protec- tive ferments in, 496 542 INDEX OF SUBJECTS Pro-agglutinoid phenome- non in agglutina- tion explained as protective colloid action, 236 Pro-agglutinoid zone in agglutination, 162 complement deviation re- action analogous to, 162 Pro-agglutinoids, 235 Prophylactic immuniza- tion, active, in man, 481 against cholera, 484. See also under Cholera against plague, 486. See also under Plague against rabies, 489. See also under Rabies against small-pox, 488. See also under Small-pox against typhoid fever, 482. See also un- der Typhoid fever "Protection," 195 "Protein fever." 367 Proteins, unknown, alexin fixation test in determination o f nature of, 211 delicacy of, 213 technique of, 212 Prototoxoids, 115 Protozoa, digestion among, and its relation to phagocytosis, 274 Ptomains, as indirect cause of infection, 30 chemistry of, 29 definition of, 31 relation of, to infection, 28 Putrefaction, chemistry of, 29 micro-organisms causing, Pyemia, 25 Rabies, active prophylactic i m m u n i zation against, 489 Hogyes method of treat- ment in, 492 Pasteur's work on, 489 preparation and attenu- ation of virus for, 490 treatment of patients in, 491 Ranzi's work on anaphy- laxis, 367 Rattlesnake poison, anti- toxin for, 466 Receptors, cell, overpro- duction of, 152 complementophile, 149 cytophile, 149 definition of, 126 of third order, 150 sessile, in anaphylaxis, 390 Resistance. See also Im- munity acquired, 60 bactericidal properties in serum and, 297 body temperature and, 51 cellular theory of, 80 Resistance, cultural condi- tions for bacteria in body and, 56 degree of phagocytosis and, 296, 297 humoral theory of, 80 increased invasive pow- ers of bacteria and, 56 individual differences and, 58 inheritance a factor in, 56 local, in skin infections, 102 natural, against infec- tion, 49 racial differences in, 55 species, to infection, 51 "Retention theory of im- munity," 83 Rhus toxicodendron. spe- cific antitoxin from, 96 Richet and He"ricourt's work on anaphy- laxis, 360 Richet and Portier's work o n anaphylaxis, 360 Richet's work on passive anaphylaxis, 380 Ricin, "protein-free," 96 Romer's method for diph- theria antitoxin standardization, 461 "Root-tubercle" bacilli, 7 Rosenau and Anderson, re- searches of, in bacterial anaphy- laxis, 410 work on anaphylaxis, 362, 374 et seq. Rosenow on variations in streptococci, 472 Roux and Yersin, experi- mental immuniza- tion in hog chol- era by, 72 Rowland's antiplague ser- um, 480 vaccine in prophylactic immunization against plague,487 Russell's vaccines for pro- phylactic immuni- zation against ty- phoid fever, 483 Salt-inactivation of alexin, 178 Salts, effect of, in agglu- tination, 243 in precipitation, 265 Saprophytes, biological transition of, to parasites, 5 occurrence of. 2 pathogenic powers of certain, 4 pure, 11 Saprophytic micro-organ- isms, definition of, 4 Scarlet fever, relative sus- ceptibility of man and animals to, 54 Schmidt, precipitation ex- periments of, on heat precipitins, 260, 261 Sensitization. 359 Bordet's views on, 162, 163 complement deviation in, 160 Ehrlich and Sachs' phe- nomenon in, 165 Bordet and Gay's in- terpretation of. 166, 167 Ehrlich and Sachs' views on, 164. 165 Neisser-Wechsberg phe- nomenon in, 160 Septicemia, chronic, adap- tation of bacteria in, 7 secondary foci in, 7 Sensibilisin, 387 Sensibilisinogen, 387 Sensitized bacteria, immu- nization with, 68 Sensitized tuberculin. 357 Sensitized vaccines, 355 Sensitizer. See Ambocep- tor Bordet's definition of, 159 quantitative determina- tion of, in im- mune serum, 160, 161 Serum. See also Blood serum antitoxic, direct effect of, on toxin. 104 indirect protective ac- tion of, against toxin, 104 bactericidal properties, of, in immunity, 81 resistance and, 297 cell-free. bacterial growth in, 81 immune, bacteriotropins in, without lysins, 322 heated, reactivation of, by addition of alexin. 318 opsonins in, bacteri- cidal sensitizers and, 321 et seq. heated, increase of power of, by ad- dition of fresh normal serum, 318 increase of, 315 normal opsonins and, 320 et seq. specificity of, 321 phagocytosis in, 91,, 315 increase of, 311. See also Phagocy- normal, agglutinins in, 91 antic omplementary properties of, 196 hemolytic properties of, 91 opsonic powers of, 314 reduction of, by heat, 314 opsonins in, 91 cooperation of heat- stable and heat sensitive body in, 318, 319 nature of, 316 INDEX OF SUBJECTS 543 Serum, normal, opsonins in, resistance of, to heat, 315 similarity of. to alexin, 316, 317 specificity of, 318 thermostability of, 320 specific thermostable opsonins in, 317 normal antitoxic, diph- theria, 107 opsonins in normal and immune, qualita- tive differences in, 315, 316 Serum sickness, 426 analogy of anaphylaxis with, 428 antibody formation in, 429 incubation time in, 429 methods of administra- tion of antitoxin to avoid, 430 Besredka's method of, 431 by alteration of serum, 430 Friedberger andMita's method of, 432 in animal experimen- tation, 431 with concentrated an- titoxin, 430 von Pirquet and Schick's studies of, 427 et seg. symptoms of, 426 accelerated reaction of von Pirquet and Schick in, 427 after first injection, 426 after second injection, 427 immediate reaction in, 427 analogy of, to ana- phylaxis, 427 Serum therapy, anaphylax- is in. See Serum sickness in diphtheria, 446. See also Diphtheria antitoxin in diseases caused by bacteria which do not form soluble toxins, 466 action of serum upon extensive infec- tion in, 467 antibacterial action in, 466 in epidemic cerebrospinal meningitis, 469. See also under C e r e b r o spinal meningitis, epi- demic in plague, 478. See also Plague, serum therapy of in pneumonia, 474. See also Pneumonia, serum therapy in In streptococcus infec- tions, 471. See also under Strep- tococcus infec- tions Serum therapy, in typhoid fever, 475. See also Typhoid fever, serum ther- apy of Side chain theory, anti- body production in body cells in, 130 body cell in, 125 chemical nature of, 126 "Leistungskern" in, 126 side chains or recep- tors in, 126 chemical action of anti- gens in, 128 definition of side chains in, 126 diagram showing cell- receptors and im- mune bodies (Ehr- lich) in, 127 in toxin-antitoxin reac- tion, 124 overproduction of recep- tors in, 129 flow of, into blood a cause of immu- nity, 128 physical mechanism of, 124 recapitulation of, 130 relationship between susceptibility p f tissue and toxin- binding properties in, 131 Skin, defence of, in infec- tion, 12 Skin infections, local im- munity in, 102 Small-pox, active prophy- lactic immuniza- tion against, 488 Jenner's discovery of, 488 production and prep- aration of vac- cine for, 488 history of experimenta- tion in immuniza- tion against, by Jenner, 62 relative susceptibility of man and animals to, 54 vaccination for, history of, 62 principles of, 62 Smith, Theobald, investiga- tions of, in ana- phylaxis, 363 phenomenon of, in ana- phylaxis, 361 Snake venoms, 36 action of. 465 activation of, by endo- complement in blood cells, 174 by sera, 174 antitoxins for, 464 effect of heat on, 105 immunization against, 464 Kyes' experiments in, 174, 175 neurotoxins in, 465 peculiarities of, 464 toxin-antitoxin combina- tion in, stability of. 105 Snake venoms, toxin-HCl modification ' of, effect of heat on, 106 toxin-antitoxin reaction with, 105, 465 filtration experiments in, 105 neutralization theory of, 105 time element in, 105 Sols, 500 Species resistance, 51 Specificity, definition of, 76 Spermatotoxins, 92 Spleen, removal of, and antibody - forma- tion in active im- munity, 100 and susceptibility to infection, 101 Standardization of anti- toxin, guinea pigs used in, 108 minimum lethal dose in, 109 Standardization of diph- theria antitoxin, by means of tox- in, 107 early attempts at, 107 unit in, 107 Standardization of sera, Pfeiffer's method of, 139 Standardization of tetanus antitoxin by means of toxin, 107 Standardization of vac- cines, 352 Hopkins' method of, 353 Wright's method of, 352, 353 Staphylococcus furunculo- sis, opsonic index in vaccine treat- ment of, 335 Staphylococcus infections, opsonic index in, 334 during vaccine treat- ment with dead s t a p h ylococcus cultures, 335 relative susceptibility of man and animals to, 54 Stern's modification o f Wassermann test, 209 Stimulins, 311 Strauss test, 25 Streptococci, variations in, 472 Streptococcus infections, agglutination re- action in diag- nosis of, 222 relative, of man and animals, 54 serum therapy in, 471 difliculties in, owing to variations of streptococci, 471, 472 early investigations in, 472 Marmorek's work on, 472, 473 nature of action in, 473 544 INDEX OF SUBJECTS Streptococcus infections, serum therapy in, standardization of serum in, 474 value of. 473 Strong's investigations in prophylactic im- munization against plague, 487 method of prophylactic vaccination i n cholera, 486 Sub-infection, 24 "Summation of negative phase" in active i m m u n i zation 338, 339 "Summation of positive phase" in active immunization, 339 Surface tension in chemo- taxis, 293 Susceptibility, body tem- perature and, 51 cultural conditions for bacteria in body and, 56 increased invasive pow- ers of bacteria and, 56 individual differences and, 58 inheritance a factor in, 56 racial differences in, 55 relative, of animals to infection, 51 species resistance to in- fection and, 51 Sycosis, opsonic index in vaccine treatment of, 336, 337 Syntoxoids, 116 Syphilis, diagnosis of, B o r d e t-Gengou phenomenon in, 188 by a 1 e x i n fixation, 198, 199 by direct precipita- tion of syphilitic serum by emul- sions of lecithin and of sodium glycocholate, 204 relative susceptibility of man and animals to, 52 ultramicroscopic method of finding precipi- tates in sera of, 204 Wassermann reaction in, diagnostic value of, 210, 211 technique of, 207 Bauer's modification of, 209 Noguchi's modifica- tion of, 208.209 schematic presen- tation of, 209 Stern's modification of, 209 Temperature, body, and re- sistance to infec- tion. 51 Tetanus, "cryptogenic," 5 relative susceptibility of man and animals to, 53 Tetanus, toxin fixation in, by brain tissues, 131 lipoidal, substances a factor in, 133 proteolytic enzymes a factor in, 133 temperature a fac- tor in, 132 Tetanus antitoxin, produc- tion of, 463 standardization of. 463 by means of toxin, 107 Tetanus bacillus, action of, 4, 5 Tetanus toxin, action of, 41 Thread reaction of Pfaund- ler in agglutina- tion, 222, 223 Thymotoxin, 92 Thyroid gland, production of alexin in, 172 production of opsonins in, 173 Toxemia, 10 Toxicity, definition of, 11 Toxin-antitoxin, diphthe- ria. See Diph- theria toxin-anti- toxin Toxin-antitoxin combina- tion, chemical re- lations of, 114 effect of heat on, 106 in snake venom, stabil- ity of, 105 toxin-HCl modification of, 106 effect of heat on, 106 measurement of, by par- tial absorption method of Ehr- lich, 115 stability of, 105 valency of component parts of, 114 Toxin-antitoxin reaction, analogy between chemical reac- tions and, 118, 119 antibody production in body cells in, 130 body cell in, 125 chemical nature of, 126 chemical action of anti- gens in, 128 concentration of reagents in, 107 degrees of toxicity in, 123 direct neutralization the protective power of, 124 effect of heat on, 105 effect of temperature on, 104 mechanism of, 104 neutralization in, absorp- tion theory of, 123 Arrhenius and Madsen on, 120 Bordet on, 122 Bordet - Danysz phe- nomenon in, 123 von Dungern's views on, 124 Danysz effect in, 123 phenomena of, 119 et seq. Toxin-antitoxin reaction, overproduction of receptors in, 128 flow of. into blood a cause of immu- nity, 128 physical mechanism of, 124 quantitative relations in, 106, 123 relationship between sus- ceptibility of tis- sue and toxin- binding properties in, 131 side chain theory in, 124 specificity of, 124, 129 speed of action of, 107 time element in, 105 with snake venom, 105 filtration experiments in, 105 neutralization theory of, 105 time element in, 105 Toxin, bacterial. See Bac- terial toxins, chemical relations of, with antitoxin, 114 deterioration theory of, 110 definition of, 32 differences in combining avidity of, 111 diphtheria, construction of, 118 normal, 107 direct effect of antitoxic serum on. 104 epitoxoid form of, 112 indirect effect of anti- toxic serum on, 104 structure of, 110 toxoid, 110 prototoxoids in, 115 syntoxoids in, 116 toxon, 113 true, 33 analogy of, with en- zymes, 36 bacteria producing, 34 characteristics of, 34 chemically indefinable nature of, 35 diseases for which some investigators claim, 469 heat sensitiveness of, 36 incubation time of, 36 production of anti- toxin by, 35 Toxin hypersusceptibility, anaphylaxis and, 407 Toxin spectra, construc- tion of, 116 definition of, 116 measurement o f, 116, 117 principles of, 116 Toxin unit, definition of, 109 diphtheria, 107 Toxoids, definition of, 110 Toxons, action of, 113 definition of, 113 structure of, 113 Toxophore group of toxin, 110 action of, 110 INDEX OF SUBJECTS 545 Tubercle bacilli, effect of body temperature on virulence of, 12 Tuberculosis, avian type, relative suscepti- bility of animals to, 52 bovine type, relative sus- ceptibility of man and animals to, 52 human type, relative susceptibility of man and animals to, 52 meistagmin reaction in diagnosis of, 497 of cold-blooded animals, immunity of warm-blooded ani- mals to, 52 opsonic index in, 341- 342, 343 Tuberculin ophthalmoreac- tion, anaphylactic nature of, 440 Tuberculin reaction, anal- ogy of, to ana- phylaxis, 442 rlactic i anaphylactic nature of, 438 Bail's experiments with passive sensitiza- tion in, 443 diagnostic value of, 442 Koch's experiments in, 439 nature of, 438 Babes' interpretation of, 439 Koch's interpretation of, 439 Wassermann and Brucks' interpre- tation of, 439 specific antibody forma- tion in, 442 Tuberculin skin reaction, anaphylactic na- ture of, 440 von Pirquet's interpreta- tion of, 441 Tuberculins, 355 Bouillon^ Filtre (Denys), 357 New Tuberculin (TR and TO). 356 New Tuberculin Ba cil- iary Emulsion, 357 Old Tuberculin (Koch), 355 Sensitized Tuberculin, 357 Tumors, malignant, alexin fixation in diag- nosis of, 213 von Dungern's method of, 214 antigen production for, 214 results of, 215 technique of, 214 et seq. organ-specific qualities in, 373 Typhoid bacilli, attenua- tion of virulence of, 18 Typhoid carriers, 3 Typhoid fever, adaptation of bacteria in, 8 agglutination reaction for diagnosis of, 219 Typhoid fever, agglutina- tion reaction for diagnosis of mac- roscopic method, 219 microscopic method, 220 effect of path of intro- duction of bac- teria of, on infec- tion, 14 meistagmin reaction in diagnosis of, 497 prophylactic immuniza- tion against, ac- tive, 482 early experimentation in, 482 living sensitized vac- cines used in, 484 results of, in United States Army, 483 Russell's vaccines in, 483 sensitized killed vac- cines in, 484 relative susceptibility of man and animals to, 54 serum therapy in, 475 Besredka's anti-endo- toxic serum in, 476 Chantemesse's early experiments in, 475 Garbat and Meyer's work on. 477 Kraus and Stenitzer's serum in, 477 Gottstein - M a t h e s' work on, 477 Liidke's work on, 477 nature of reaction in, 476 Typhus fever, relative sus- ceptibility of man and animals to, 54 Ultramicroscope, 503 Vaccination, prophylactic, in man, 481 in anthrax, 64 in cholera, 484. See also under Cholera in plague, 486. See also under Plague in rabies, 489. See also under Rabies in small-pox, 488. See also under Small- pox history and general principles of, 62 in typhoid fever, 482. See also under Typhoid fever Vaccine therapy. See also Immunization, ac- tive anaphylaxis in, 432 as a therapeutic meas- ure, action of, in local infections, 346 in generalized sys- temic infections, 347 in successive local infections, 347 value of, in acute dis- eases, 350 Vaccine therapy, as a therapeutic meas- ure, value of, in subacute or chronic cases, 349 autoinoculations by mas- sage or exercise in, 340 "high tide" of immunity in, 340 "negative" phase in, 338 second injection in, 338 successive inoculations in, 338, 339 summation of, 338. 339 opsonic index in, 328 et seq. comparison between that in exudate of infected foci and blood serum, 340 improvement and, 341 in tuberculosis, 341- 343 Leishmann's technique for determination of, 329 Simon, Lamar and Bispham's tech- nique for deter- mination of, 332, 333 value of. 338 in controlling thera- peutic vaccina- tions, 344 in showing degree and conditions in which vaccination is successful, 344, 345 Wright's technique for determination of, 330 et seq. difficulties in, 332, 333 value of, 333 relation of phagocytosis to, 329 second positive phase in, 339 "summation of positive phase" in. 339 tuberculins in, 355 value of, as prophylactic measure, 345, 346 as therapeutic meas- ure, 346 Vaccines, autogenous, 351 production of, 351 with dead bacteria, 351 with living bacteria, 351 sensitized, 355 standardization of, 352 Hopkins' method of, 353 Wright's method of, 352, 353 Vaughan's work on ana- phylaxis, 366.367 • on bacterial anaphylaxis, 412 Vaughan and Wheeler, pro- teid split prod- ucts of, in ana- phylactic poison, 403 theory of, on mechanism of anaphylaxis, 393 546 INDEX OF SUBJECTS V a u g h a n and Wheeler, work of, on toxic fraction of pro- tein molecule in anaphylaxis, 393 Virulence, aggressin secre- tion of bacteria in body and, 20-22 capsule formation of bacteria and, 18 definition of, 11 dependent on resistance of bacteria to leu- kocytes in pha- gocytosis, 325 ectoplasmic hypertrophy of bacteria in re- lation to, 19, 20 effect of body tempera- ture on, 12 effect of cultural adap- tation of bacteria on, 12 effect of path of intro- duction of bac- teria on, 12-14 effect of quantity of bac- teria introduced on, 14 increase of, by attenua- tion of bacteria, 17 measurement of relative deg ees of, 15 of capsulated bacteria, 326 relation of, to phagocy- tosis, 312 relative to number of bacteria intro- duced, 15 specificity of bacteria and, 22 variation in, of bacteria successiv ely passed through animals, 16. 17 of different strains of same bacteria, 15, 16 Virulins, 22, 326 "Virus fixe" in treatment rabies, 490 Wassermann reaction, 198 alexin fixation principle in, 198 not by union of spe- cific syphilitic an- tigen with spiro- chseta pallida an- tibodies, 204 alexin titration in, 206 antigen preparation for, 200 by addition of choles- trin, 201 by method of Brown- ing and Cruik- shank, 201 by method of Noguchi, 200 by methods of Forges and Meier, 200 by methods of Weil and Braun. 200 titration in, 202 diagnostic value of, 210, 211 in diagnosis of syphilis, 198 in diseases other than syphilis, 210 in normal organs, 200 Klausner theory in, 204 precipitation in, by addi- tion of syphilitic serum to lecithin emulsions, 204 produced with syphilitic serum in antigens from normal or- gans, 200 specific antigen from spirochaeta pallida cultures unsuit- able in, 203 Wassermann reaction, spinal fluid used in performance of, 210 technique of perform- ance of, 207 Bauer's modification of, 209 Noguchi's modification of, 208, 209 schematic presenta- tion of, 209 refrigerator method in, 208 Stern's modification of, 209 schematic presentation of, 207 theories of, 204, 205 titration of hemolytic amboceptor o r sensitizer in, 205 ultramicroscopic method of finding precip- itates in syphili- tic sera in, 204 Weigert's law of overcom- pensation. 128 Wright's method of stand- ardization of vac- cines. 352, 353 Wright's technique for de- termination of op- sonic index, 330 et seq. difficulties in, 332, 333 value of, 333 Wright's studies of bac- tericidal and ag- glutinating pow- ers of blood se- rum, 328 Yellow fever, susceptibility to, 55 Yersin anti-plague serum, 478-480 XJZ RETURN BIOSCIENCE & NATURAL RESOURCES LIBRARY TO — *• 2101 VALLEY LIFE SCIENCES BLDG. 642-2531 LOAN PERIOD 1 ** ^ g f* p r 2 -• • : - | 3 ^!AM 4U^£ 111 W** I H L uftW ALL BOOKS MAY BE RECALLED AFTER 7 DAYS DUE AS STAMPED BELOW . /\ j \fi 1 f^ AJ K ^ T~ I UNIVERSITY OF CAUFORNIA, BERKELEY FORM NO. DDO, 50m, 1 1/94 BERKELEY, CA 94720 2343! u'c'BERimi« CD5ED7E13b is i Z - UNIVERSITY OF CALIFORNIA LIBRARY