STUDIES ON THE MECHANISM OF RECOVERY IN PNEUMOCOCCAL PNEUMONIA : I. THE ACTION OF TYPE SPECIFIC ANTIBODY UPON THE PULMONARY LESION OF EXPERIMENTAL PNEUMONIA

A uniformly fatal lobar pneumonia was produced in white rats by inoculation of the left main bronchus with virulent Type I pneumococci suspended in mucin. All of the animals succumbed in less than 5 days, half of them dying within 48 hours. In only 5 of 40 rats was the lesion confined to the left lung, and all but one developed pleurisy, pericarditis, or both. All had bacteriemia at the time of death. The pathogenesis of the pulmonary lesion was studied by examining the lungs of 35 rats killed at various intervals following inoculation. The pneumonic process spread rapidly until most of the left lung was involved in 36 hours. Frequent blood cultures showed invasion of the blood stream in a few rats at 6 hours and in over 90 per cent at the end of the first day. The first signs of pleurisy usually appeared at 18 hours. Microscopic examination of the actively spreading lesion revealed three characteristic zones: (1) an outer "edema zone" in which the alveoli contained many pneumococci floating freely in edema fluid, (2) a middle zone where both leucocytes and organisms were present, many of the latter being phagocytized, and (3) an inner zone of advanced consolidation in which the alveoli contained many leucocytes but no organisms and where there were already local areas of early resolution. Study of numerous lesions, at intervals of from 12 to 36 hours after inoculation, indicated that the pneumococci spread into normal alveoli principally by way of the infected edema fluid in the outer zone. Pneumococcus-laden edema fluid in large bronchi and in alveoli beneath the pleura suggested the mode of spread of the infection to other lobes and possibly to the pleural cavity. No adequate explanation could be found for the presence of active phagocytosis in the lungs of animals with bacteriemia and presumably without circulating antibodies, but this conspicuous phagocytic reaction was obviously responsible for the clearing of the central part of the spreading lesion. The action of type specific antibody upon the pulmonary lesion of experimental lobar pneumonia was studied in rats similarly infected but treated with antipneumococcal serum. When injected intravenously in a single dose within 18 hours after inoculation the antiserum was found to protect all of the rats against the otherwise fatal pneumonia. It stopped the spread of the pneumonic lesion, cleared the blood stream of organisms, and prevented the extension of early pleurisy. The antibody caused agglutination and capsular swelling of the pneumococci in the lung, particularly in the edema zone at the margin of the lesion where they were most numerous. Apparently immobilized by agglutination the organisms were overtaken by leucocytes and destroyed by phagocytosis. The phagocytic reaction was greatly accelerated by the specific opsonins of the antiserum, and the pneumococci were destroyed by polymorphonuclear leucocytes before many macrophages appeared in the alveolar exudate. Within a week after treatment resolution of the pulmonary lesion was well in progress. Both horse and rabbit antibody were shown to penetrate the lung, and immune bodies were demonstrated in the alveoli within 10 minutes after the start of treatment. The relation of the observed phenomena to the curative action of anti-pneumococcal serum has been briefly discussed, and it is pointed out that the principal effect of antiserum is to cause immobilization of the pneumococci in the advancing edema zone. Experiments to be reported in a later publication have shown that sulfapyridine exerts a similar effect through a different mechanism.     

EXPERIMENTAL PNEUMONIA IN GUINEA PIGS : I. THE EFFECT OF CERTAIN TOXIC AUTOLYSATES OF PNEUMOCOCCI.

1. Anaerobic autolysates of pneumococci, prepared according to the method described, are highly toxic for guinea pigs when injected intratracheally in dosage of 0.2 cc. Death occurs either within a few hours (36 per cent) or within 3 days. In the early deaths there is intense hemorrhagic edema of the lungs with beginning inflammatory reaction; in animals surviving for 18 hours or longer extensive areas of pneumonia are produced. 2. The intratracheal injection of virulent living pneumococci is followed by transient slight lesion with recovery, or by later death from septicemia without pneumonic lesions. 3. The addition of a sublethal dose of toxic autolysate to living pneumococci alters the reaction of the animal, so that there develops extensive pneumonia associated with unrestrained multiplication of the organism.     

PULMONARY EDEMA IN INFLUENZAL PNEUMONIA OF THE MOUSE AND THE RELATION OF FLUID IN THE LUNG TO THE INCEPTION OF PNEUMOCOCCAL PNEUMONIA

Pulmonary edema is a component of the fully developed influenza viral lesion in the mouse. Mice with experimental pulmonary fluid have an increased susceptibility to inhaled pneumococci and under these circumstances the organisms grow in the lung and produce the lesion of bacterial pneumonia. The presence of pulmonary edema in the lesion due to the influenza virus in the lung of the mouse appears to account adequately for the previous observation that inhaled pneumococci grow in the influenza viral lesion. Mice dying of pneumococcal septicemia after inhaling fine droplets containing this organism do not have pneumonia. The delay in migration of polymorphonuclear leucocytes into the lung after injection of pneumococci suspended in serum is an important factor in susceptibility to infection since it allows ample time for pneumococci to grow in the pulmonary fluid. The slow phagocytic action of pulmonary macrophages likewise permits growth of pneumococci. Conditions in human beings that are known to be complicated by pulmonary edema are also known to be associated with increased susceptibility to secondary bacterial pneumonia.     

STUDIES ON THE MECHANISM OF RECOVERY IN PNEUMONIA DUE TO FRIEDL?NDER'S BACILLUS : II. THE EFFECT OF SULFONAMIDE CHEMOTHERAPY UPON THE PULMONARY LESION OF EXPERIMENTAL FRIEDL?NDER'S BACILLUS PNEUMONIA

Sulfonamide chemotherapy was found to cure rats of an otherwise fatal form of experimental Friedländer''s bacillus pneumonia when treatment was begun 6 hours after inoculation. Most of the pneumonic lesions cleared completely, but an occasional animal exhibited small residual abscesses in the previously consolidated lung. The recovery process taking place in the lungs was studied histologically at various intervals during therapy. As in the case of pneumococcal pneumonia, the principal action of the sulfonamide was upon the bacteria in the advancing edema zone at the periphery of the pneumonic lesion. The bacteriostatic action of the drug appeared to stop the spread of the pneumonia, and the Friedländer bacilli were ultimately ingested and destroyed by the phagocytic cells in the alveolar exudate. The phagocytosis of bacteria in the lung was shown to be unrelated to the presence of antibody in the blood.     

STUDIES ON THE MECHANISM OF RECOVERY IN PNEUMOCOCCAL PNEUMONIA : IV. THE MECHANISM OF PHAGOCYTOSIS IN THE ABSENCE OF ANTIBODY

1. Evidence has been presented in previous publications that the phagocytosis of pneumococci in the pneumonic lung during chemotherapy is due neither to specific opsonins nor to capsular injury (1, 2). The present studies have shown that the phagocytosis taking place in the lung is independent of any sort of intermediary factor and results from a direct action of the phagocytic cells upon the pneumococci. 2. Phagocytosis in the absence of antibody has been demonstrated not only in the lungs of living rats but in formalin-fixed lungs, on the surfaces of a variety of tissues (both freshly removed from the animal and previously "killed" with heat), and on the surfaces of such inert materials as moistened filter paper, cloth, and fiber glass. On the other hand, smooth materials such as glass, cellophane, albumin, and paraffin have failed to support the phagocytic reaction. This latter observation indicates that the physical character of the surface to which the leucocytes have access constitutes a determining factor in the non-antibody mechanism of phagocytosis. 3. Further experiments have defined the relationship of "surface phagocytosis" to that induced by specific opsonins. The non-antibody mechanism was found to operate only upon surfaces of suitable physical properties, whereas opsonins enabled phagocytes floating freely in a fluid medium to engulf the fully encapsulated organisms. 4. Direct visualization of the surface phenomenon in the lung revealed that leucocytes phagocyte the virulent organisms in the absence of antibody only after having trapped them against the alveolar walls. Once the encapsulated pneumococci have been ingested, they can be seen to undergo digestion within a few hours. The discovery of the phenomenon of surface phagocytosis affords clarification of previously unanswered problems concerning the mechanism of recovery in pneumococcal pneumonia.     

STUDIES ON THE MECHANISM OF RECOVERY IN PNEUMOCOCCAL PNEUMONIA : III. FACTORS INFLUENCING THE PHAGOCYTOSIS OF PNEUMOCOCCI IN THE LUNG DURING SULFONAMIDE THERAPY

The phagocytosis which occurs in the lungs of rats receiving sulfonamide is due neither to an opsonizing action of the sulfonamides nor to type-specific antibody. The evidence presented indicates that the destruction of the pneumococci is brought about by a phagocytic mechanism independent of both opsonization and capsular injury.     

THE HISTOGENESIS OF CELLS IN EXPERIMENTAL PNEUMONIA IN THE DOG

The kinds of cells and the sequence in which they appeared in the inflammatory exudate were studied in a series of experimentally produced pneumonic lesions in dogs. There was a gradual and progressive change in the character of the exudate and the kinds of cells as the disease progressed. The microscopic findings could be more closely correlated with the age of the lesions than with their gross appearance or with the clinical condition of the animal at the time of death. The cells in the exudate came principally from the blood. The polymorphonuclear leucocytes were gradually replaced by larger phagocytic mononuclear cells. These were derived chiefly from the hypertrophy and transformation into larger phagocytic cells of the lymphocytes and monocytes of the blood after they entered the air spaces along with the polymorphonuclear leucocytes in the early stages of the disease. To follow the development of the hematogenous exudate cells into macrophages in the dog, the pneumonic process must be studied from its earliest inception and at close stages during the first 36 hours of the disease. The local septal cells contributed only in a minor way to the origin of the macrophages. Their principal reaction appeared to be one of enlargement without detachment from the alveolar walls. Consolidation of the lungs occurred as a result of the spread of the pneumococci through the air spaces by direct passage of the infected edema fluid from alveolus to alveolus through the pores of Kohn and from bronchiole to bronchiole from aspiration during breathing and coughing. The similarity of the histogenesis of the exudate cells in this series of experimentally induced pneumonic lesions in the dog and in those which occur spontaneously in man was discussed.     

A STUDY OF THE MACROPHAGE REACTION IN THE PULMONARY LESIONS OF DOGS WITH EXPERIMENTAL PNEUMOCOCCUS LOBAR PNEUMONIA

A study has been made of the occurrence of the macrophage reaction in the pulmonary lesions of dogs sacrificed during the course of experimentally induced pneumococcus lobar pneumonia or dying as a result of the infection. This characteristic transformation of the fixed tissue cells of the lung was found as a constant accompaniment of recovery. It was also present in varying degrees in the great majority of fatal instances provided the animal lived more than 40 hours. In general the longer the animal survived, the more pronounced the macrophage reaction observed in the lung lesions at autopsy. The numbers of pneumococci in the lesions diminished progressively with the evolution of the cellular change which terminated in resolution of the pneumonic exudate. Some dogs surviving for 4 days or more showed practically complete clearing of the pulmonary lesions but succumbed with an overwhelming bacteremia or empyema or both. On the other hand, several animals dying with a sterile blood, exhibited lesions characterized by little or no macrophage response and the presence of many pneumococci. These findings suggest that recovery from experimental lobar pneumonia in the dog depends on a dual mechanism consisting of a generalized process which prevents or controls invasion of the blood stream and a local one by which the lesion is finally freed from the invading microorganisms. The nature of these two processes is discussed.     

THE RELATION OF NATURAL HUMORAL ANTIPNEUMOCOCCAL IMMUNITY TO THE INCEPTION OF LOBAR PNEUMONIA

A study of the pneumococcidal-promoting action of the serum of lobar pneumonia patients, secured from 4 to 48 hours after the onset of the disease, has revealed the fact that in the majority of instances the serum possessed the power to promote killing of the homologous pneumococcus, isolated in different instances from the lung, blood, and sputum. While in some instances this action was slight, in others it was present to as great a degree as in normal individuals and persisted as long as 48 hours or more after the beginning of the disease. The variations observed from case to case were not related to the extent of the pneumonic lesion or to the virulence of the several pneumococcus strains but appeared to depend on differences in individual human beings in respect to the natural antipneumococcus properties of their blood and their reaction to the invading microorganism. A constant relationship was found to exist between the concentration of immune properties in the serum and blood invasion. In the presence of a well marked pneumococcidal-promoting power pneumococci were not found in the blood stream, and only when this property was greatly diminished or lost did blood invasion occur. The findings which are supported by certain previous experimental observations, indicate that lobar pneumonia can occur in the presence of a normal circulating antipneumococcus defense mechanism. From this it is inferred that before pneumococcus growth can be initiated there must be present in the lung local changes of such nature as to provide conditions for the multiplication of pneumococci protected from the pneumococcidal action of the blood. Suppositions as to the nature of these changes and the establishment of the pneumonic lesion are discussed.     

STUDIES ON THE PATHOGENESIS OF EXPERIMENTAL PNEUMOCOCCUS PNEUMONIA IN THE DOG : I. SECONDARY PULMONARY LESIONS. RELATIONSHIP OF BRONCHIAL OBSTRUCTION AND DISTRIBUTION OF PNEUMOCOCCI TO THEIR INCEPTION

A study has been made of the mode of origin of the secondary lesions occurring spontaneously during the course of experimental pneumococcus pneumonia in the dog. It was observed that the primary lesions of dogs exhibiting interlobar spread contained much more edema fluid than did those in which the inflammatory process remained confined to a single lobe. Furthermore, the sequence of spread from lobe to lobe in relation to the anatomical arrangement of the bronchi and the prone position of the animal was such as to suggest that secondary lobe involvement arose as a result of edematous pneumonic exudate flowing into the more dependent bronchial openings. Experiments were undertaken to determine whether pneumococci are constantly being distributed throughout the lung in the experimental disease in varying degree yet produce secondary foci of consolidation only if the microorganisms reach the terminal airways and are retained there through some interference with the normal eliminatory mechanism. Attempts to produce secondary lesions in dogs with non-spreading single lobe involvement, by means of plugging a terminal bronchus of a normal lobe with starch paste or mucin were largely unsuccessful. In only three out of 19 instances did a lesion develop at the site of obstruction. An investigation was then made of the distribution of pneumococci in the lungs of dogs at the height of the pneumonia. In dogs with single lobe lesions pneumococci were recovered from the lesion itself but not from any other part of the peripheral lung tissue, whereas in animals showing spread to other lobes pneumococci were found to be distributed widely throughout the lung in both the apparently normal and the involved lobes. Some of the microscopic sections of the uninvolved parts of the lungs of dogs with metastatic lesions revealed small masses of pneumococcus-containing exudate in the smaller bronchi and terminal airways of otherwise normal tissue. This finding, in conjunction with the detection of beginning inflammatory changes in other areas normal in the gross, would seem to provide direct evidence of the manner in which pneumococci are transferred from the initial lesion to other lobes and highly presumptive evidence for the mode of origin of the secondary lesions.     

LOCAL RECOVERY IN EXPERIMENTAL PNEUMOCOCCUS LOBAR PNEUMONIA IN THE DOG

An investigation has been made of the process of local recovery occurring in the pulmonary lesions of dogs with experimental pneumococcus lobar pneumonia. Six animals showing simultaneous healing and spread of the pathological process in different parts of the lungs were sacrificed during the active stage of the disease for bacteriological and histological study. It was found that with few exceptions the clearing (resolving) lesions as revealed by x-ray during life were sterile on culture while the young, metastatic processes yielded an abundant growth of pneumococci. The resolving areas which represented the older lesions were characterized by the presence of a well developed macrophage reaction, whereas the early lesions consisted of alternating areas of edema and polymorphonuclear infiltration and contained usually many pneumococci both free and in the cells. The only pneumococci found in the areas of macrophage mobilization were intracellular forms, for the most part in the process of digestion. In several instances these striking differences in the cellular picture and the numbers and distribution of pneumococci were observed in different parts of a single lobe lesion. Tests on the blood serum failed to reveal any evidence of acquired antipneumococcal humoral immunity. The significance of these findings in relation to the mechanism of recovery is discussed.     

THE PATHOGENESIS AND PATHOLOGY OF EXPERIMENTAL TYPE I PNEUMOCOCCIC PNEUMONIA IN THE MONKEY

The pathogenesis of lobar consolidation and microscopic pathology of induced Type I pneumococcic pneumonia in a series of fourteen monkeys, killed at close intervals of time after infection, have been studied. The inflammatory process which resulted in consolidation was primarily intraalveolar and intrabronchial. The pneumococci spread within the air spaces as a result of the dissemination of the infected edema fluid directly from alveolus to alveolus through the pores of Kohn and from bronchiole to bronchiole as a result of repeated aspiration during breathing. The pneumonic process within the air spaces developed and progressed independently of the reaction in the interstitial tissues. The organisms spread to the interstitial tissue secondarily from the alveolar spaces. Once in the interstitial tissues they appeared to be the important source of infection producing bacteremia but not important in the mechanism by which consolidation was produced. The exudate cells came chiefly from the blood. The large mononuclear cells which replaced the polymorphonuclear leucocytes were derived principally from the hypertrophy and transformation of lymphocytes and monocytes into macrophages after they entered the exudate in the early stages of the disease. The part the local septal cells played as the source of the macrophages could not be accurately determined. The reaction of the septal cells appeared to be chiefly one of swelling without detachment and occasional proliferation to form binucleated attached cells. To follow the transformation of the hematogenous mononuclear cells into macrophages in the exudate, the inflammatory reaction must be examined at frequent intervals during the first 36 hours of the disease. The similarity of the pathogenesis of lobar consolidation in human pneumonia to that observed in the experimentally induced disease in monkeys and dogs was discussed.     

STUDIES ON ENDEMIC PNEUMONIA OF THE ALBINO RAT : III. CARRIAGE OF THE VIRUS-LIKE AGENT BY YOUNG RATS AND IN RELATION TO SUSCEPTIBILITY

An attempt was made to reestablish the virus-like agent associated with rat pneumonia by the nasal instillation of the infectious material into supposedly normal young rats from the selected Princeton colony. The incidence of the pneumonia in these animals after the injection of suspensions of pneumonic lungs from either rats or mice was not significantly greater than the incidence under natural conditions. In attempting to account for the refractory state of the immature rats it was found that the agent was widely dispersed through the breeding colony at an early age. Detailed tests of many litters indicated that the agent was acquired shortly after birth by way of the upper air passages as the result of maternal contact. In most instances it was so well tolerated by young rats during the suckling period and for several months thereafter that its presence was recognized only by nasal transfer to mice. There is every indication that the rate of infection of rats in the breeding colony approaches 100 per cent by the time the animals are old enough to be used experimentally.     

EFFECT OF PNEUMOCOCCUS AUTOLYSATE ANTITOXIN ON PNEUMOCOCCUS PNEUMONIA IN GUINEA PIGS

1. In rabbits, guinea pigs, and mice pneumococcus autolysate antitoxin has no effect against generalized infection by the pneumococcus. 2. In a large proportion of cases of pneumonia and sepsis caused by the intratracheal inoculation of large amounts of living pneumococci in normal or sensitized guinea pigs, autolysate antitoxin saves the animal. 3. The autolysate antitoxin is protective in most cases of pneumonia and sepsis caused by the intratracheal inoculation of small amounts of living pneumococci combined with sublethal doses of staphylococcus toxin. 4. These experiments indicate that in pneumococcus pneumonia of guinea pigs the autolysate toxin or toxins are elaborated only in the lungs.     

THE ACTION OF THE LETHAL DOSE OF STROPHANTHIN IN NORMAL ANIMALS AND IN ANIMALS INFECTED WITH PNEUMONIA

The results of these experiments permit a comparison of the action of strophanthin in the normal and the infected animals. In cats the percentage of deaths following the injection of 0.1 mg. of strophanthin per kilo of body weight was the same in normal and in pneumonic cats. The number of recoveries in this series was larger than was expected. This result was due to the fact that the dose injected was not the lethal dose, but the average lethal dose (0.1 mg.) determined by Hatcher and Brody (7) and by Eggleston (8). The doses which the latter actually injected ranged from 0.085 rng. to 0.16 mg.; from these the average minimal lethal dose was calculated. In adopting the average dose as the standard one to inject, those of our cats that required more than 0.1 mg. naturally survived. The death rate was therefore low. The plan employed in dogs differed from that used in cats. A lethal dose was injected in each dog. Death occurred when an average of 0.12 mg. of strophanthin per kilo of body weight was injected. The same dose was required in normal and pneumonic dogs. The effect of strophanthin in both groups of infected and non-infected cats and dogs is, therefore, identical. Whether the uniformity of strophanthin action in the two experimental groups may serve as the basis for assuming a like uniformity of action in normal individuals and in pneumonia patients is a subject which requires further analysis. The difficulty in transferring the experimental results to patients lies in the question of whether the type of pneumonia produced in animals is the same as that found in man. Clinically, the two diseases present both resemblances and differences. The animals become definitely ill and show the symptoms already described. The illness, however, is of short duration and apparently reaches its height in the majority of animals in twenty-four to seventy-two hours. Before the expiration of this time, the temperature frequently returns to normal. Many of the infected animals, when they survive, recover in three to five days. The mortality in dogs infected with pneumococci is given by Lamar and Meltzer as 16 per cent. In the present series of twelve infected cats, the mortality was also 16 per cent. These findings differ from human pneumonia in the following particulars: The infection is not so severe; the temperature, though elevated at first, soon falls; the duration of the disease is short; and convalescence is rapid. The mortality is slightly lower. Musser and Norris (12) give the human mortality at 21.06 per cent. Pathologically the two diseases also show differences. The gross appearance of the lungs is not dissimilar, but in the animals the consolidated portions are somewhat dry and they fail to show a stage of gray hepatization (5). The amount of fibrin present is small. There is comparatively slight congestion of the alveolar walls and of the walls of the bronchi. The relation of experimental pneumonias to the human disease has been discussed by a number of investigators. Almost all believe that the two types are similar, if not identical. Among the first to express this opinion was Sternberg (13) ; and later Gamaleia (14), Prudden and Northrup (15), Kinyoun and Rosenau (16), Wadsworth (17), Lamar and Meltzer (5), Wollstein and Meltzer (9) coincided with his view. Lamar and Meltzer, especially, have insisted on the identity of the two processes. On the other hand, Welch (18) in his study of experimental pneumonia, says: "Many inoculations of cultures of virulent pneumococci into the trachea and lungs of dogs have been made in my laboratory by Dr. Canfield and myself, but in no instance were we able to produce an inflammation of the lungs which we were willing to identify with acute lobar pneumonia as found in human beings." But he adds that, in the majority of experiments, there was no demonstrable consolidation and that pleurisy and more or less extensive areas of pneumonia were produced only in a few animals. The inference consequently cannot be drawn that an effect obtained with strophanthin in the experimental disease may be anticipated in man. The striking similarity in action in infected and uninfected animals renders it likely, however, that the usual action of the drug in man may be expected in the presence of pneumonia. We have accumulated evidence, to be published later, which shows that this action actually takes place in the human disease. As far as evidence obtained electrocardiographically is concerned, our experiments show that strophanthin causes the same electrical changes in the heart when the animals are infected as it does under normal conditions.     

PATHOLOGY OF EXPERIMENTAL PNEUMOCOCCUS PNEUMONIA IN MICE

1. The administration of alcohol produces no evident histological changes in the lungs of mice. 2. Following inhalations of pneumococci the organisms are not visible histologically in the lungs of mice either in the bronchi or alveoli. 3. In partially immunized mice which have been exposed to a pneumococcus spray while alcoholized, true lobar pneumonia not infrequently develops. 4. The primary lesion of such lobar pneumonia in mice is an interstitial inflammation of the alveolar wall and the infection spreads in the interstitial tissue like a cellulitis. 5. A tentative explanation of the "lobar" distribution of pneumococcus pneumonia is offered.     

STUDIES ON THE MECHANISM OF RECOVERY IN PNEUMONIA DUE TO FRIEDL?NDER'S BACILLUS : I. THE PATHOGENESIS OF EXPERIMENTAL FRIEDL?NDER'S BACILLUS PNEUMONIA

Experimental pneumonia due to Friedländer''s bacillus was produced in white rats by the intrabronchial inoculation of the bacilli suspended in mucin. The pneumonia was lobar in type, was almost uniformly fatal, and simulated the acute form of the natural disease in human beings. The pathogenesis of the pneumonic lesion was studied by examination of microscopic sections of the lungs of animals killed at frequent intervals during the course of the infection. The histologic characteristics of the various stages of the pneumonia were essentially the same as those previously described in experimental pneumococcal (Type I) pneumonia except for the following differences: (1) In isolated areas of the lung in Friedländer''s pneumonia many more bacteria were encountered in the alveoli than were ever noted in experimental pneumococcal pneumonia. (2) Abscess formation was common in the late stages of Friedländer''s infection, whereas it was not noted in the pneumococcal lesion. (3) Organization of the alveolar exudate, rarely observed in experimental pneumococcal pneumonia, was a prominent feature of the pneumonia due to Friedländer''s bacillus. The mechanism of spread of Friedländer''s lesion appeared to be the same as that of pneumococcal pneumonia. Likewise there was noted the same phagocytosis of organisms in the lungs of even bacteremic animals dying of the infection.     

STUDIES UPON EXPERIMENTAL PNEUMONIA IN RABBITS : VII. THE PRODUCTION OF LOBAR PNEUMONIA.

1. The production of lobar pneumonia in rabbits is dependent upon the introduction of organisms into the alveoli themselves. 2. In order to accomplish this the catheter through which they are injected must be inserted as deeply into a bronchus as possible and the culture fluid injected with considerable force. 3. Large numbers of organisms injected into the trachea just beyond the larynx set up no great changes in the lungs, even though the injection be forcible. This fact suggests the presence of a protective mechanism in the upper air passages, which, under normal conditions, prevents the penetration of organisms into the lungs. 4. If animals be subjected to cold, alcohol, and the inhalation of irritating gases, the so called secondary factors in the etiology of lobar pneumonia in man, then the injection of pneumococci into the trachea causes inflammatory changes of the upper respiratory tract and occasionally pneumonia. 5. The vagi prevent foreign material in the pharynx and upper respiratory tract from reaching the lungs. Section of one vagus may be followed by pneumonia, while section of both invariably leads to this result. 6. It is possible that the secondary factors mentioned above owe their action to their influence upon the vagus control of the upper respiratory tract.     

THE TREATMENT OF ANOXEMIA IN PNEUMONIA IN AN OXYGEN CHAMBER

1. The use of an oxygen chamber in the treatment of pneumonia patients makes it possible to administer this gas for long periods of time under exactly known conditions. The medical and nursing care of the patient is greatly facilitated. 2. Prolonged inhalation of oxygen varying from 40 to 60 per cent appears to be without harm. 3. Oxygen administered to intensely anoxemic patients almost immediately clears up this anoxemia. Cyanosis disappears with the anoxemia. 4. The removal of patients from the high oxygen while they are still sick and while examination shows that there are still extensive edema and infiltration of the lung results in a return of the intense anoxemia. 5. It is sometimes impossible to clear up the anoxemia, even when as high as 60 per cent of oxygen is given, especially when there are considerable edema and infiltration of the lungs. 6. Five cases in which the prognosis was grave recovered. Three cases, one of tuberculosis, one with a Pneumococcus Type III infection, and a third with a pneumonia superimposed on a chronic pulmonary condition, died. 7. In all cases there appeared to be an improvement in the patient''s condition. In one case, particularly, with an intense degree of anoxemia, the patient became moribund and pulseless. Following the administration of 60 per cent of oxygen there was a lowering of the heart rate from 160 to 120, the return of the pulse to the radial artery, the color became bright pink, and there was a remarkable change in the clinical condition. 8. The anoxemia of pneumonia is due, in large measure, to an impairment of the respiratory surface of the lungs. The greater the lung involvement the greater the anoxemia. Especially is this so when the pneumonic process extends throughout the lungs so that there are many patches of bronchopneumonia, with accompanying bronchitis and edema, as evidenced by the presence of râles throughout the lungs. 9. Rapid and shallow breathing of the degree observed in pneumonia is, as far as present evidence shows, of less importance in the production of anoxemia.     

THE SPECIFIC POLYSACCHARIDE CONTENT OF PNEUMONIC LUNGS

1. The specific polysaccharide content of pneumonic lungs in Type III pneumonia was 60 times greater than in Type I, II, VII, and VIII cases. 2. The highest values were obtained from gray hepatized lobes, but the red hepatized and edematous areas also contained large quantities of S. 3. Comparable yields of polysaccharide were recovered from the sputa and lungs of fatal cases. 4. Preliminary peptic digest of lung exudates increased the S yields. 5. The data support the hypothesis that the outcome in Type III pneumonia is related to the ability of the pneumococci to produce capsular polysaccharide.