There is no reason for believing that the influenza which prevailed in this country differed in any essential feature from that of previous epidemics and particularly of the pandemic of 1889–90. Our studies have shown that an organism with the morphologic and cultural characters of B. influenzæ of Pfeiffer has been constantly found in association with the disease, and so frequently demonstrated in association with its pulmonary complications that there is little doubt of its constant presence. The bronchial and pulmonary complications of influenza present characters which, while varied, are not usually observed in the absence of epidemic influenza, and in this pandemic agree with those of the former pandemic so far as it is possible to determine from the descriptions available.

Especially noteworthy is the severity of the changes within the bronchial passages. Clinical studies have shown that purulent bronchitis has occurred in 36 per cent of instances of influenza. The sputum with this condition has contained B. influenzæ in all instances, but although there were no signs of pneumonia it has been constantly associated with other microorganisms, namely, pneumococci (in 11 of 13 instances), S. hemolyticus, S. viridans, M. catarrhalis, etc.

Identification of the bacteria which have been present in the bronchi of those dead with pneumonia following influenza have determined what microorganisms have penetrated into the lower respiratory passages. B. influenzæ has been found so frequently (80 per cent) that there is good reason to believe that it has been constantly present and has not been isolated in every instance because it has been overgrown by other microorganisms on the plates or after long continued illness has disappeared from the bronchi. Mixed infections of B. influenzæ and other microorganisms are constantly found in the inflamed bronchi; combinations of B. influenzæ and pneumococci, B. influenzæ and hemolytic streptococci or these combinations with staphylococci or the four organisms together are common. Other microorganisms such as B. coli, S. viridans, M. catarrhalis and diphtheroid bacilli are not infrequently associated with those which have been mentioned.

Purulent bronchitis has been found in 137 of 241 autopsies; its bacteriology differs in no respect from that which has just been described and indeed no line can be drawn between this condition and the bronchitis invariably present with the pneumonias of influenza. Other evidence of profound injury to the bronchi is the frequent occurrence of hemorrhage in a zone ensheathing the smaller bronchi, and the common occurrence of bronchiectasis when the fatal disease has lasted more than two or three weeks.

Microscopic study demonstrates that the changes in the bronchial walls are such as destroy the defences against invasion by microorganisms. The bronchial epithelium undergoes destruction which is not infrequently limited to the superficial ciliated cells, but often complete loss of epithelium occurs. The mucous glands of the larger bronchi exhibit a special susceptibility to injury, and in the early stages of the lesion profound degenerative changes are found in the secreting cells, whereas at a later stage chronic inflammatory changes are almost invariably present.

Pneumonia following influenza is in most instances bronchopneumonia, but typical lobar pneumonia has been found in autopsies representing 40.7 per cent of pneumonias of influenza. Lobar pneumonia is frequently accompanied by purulent bronchitis, and in a considerable number of autopsies (34 of 98 with lobar pneumonia) lobar and bronchopneumonia have occurred in the same individual.

Statistics based upon the clinical diagnosis of lobar and bronchopneumonia following influenza are so inaccurate that they have little if any value. Notwithstanding careful study of the symptomatology of the disease, lobar and bronchopneumonia following influenza are not accurately distinguishable by the means usually employed, and an erroneous diagnosis has been recorded on the patient’s history in 36.6 per cent of 227 fatal cases with autopsy. A diagnosis of suppurative pneumonia is rarely if ever made. The difficulties of diagnosis are in part explained by the frequent association of lobar pneumonia with purulent bronchitis, with bronchopneumonia or with both, and by the occurrence of bronchopneumonia with confluent lobular consolidation involving a large part of a lobe or whole lobes.

There are many defects in the present knowledge of the symptomatology of the pneumonias under consideration. The symptoms of suppurative pneumonia are not clearly defined. Many of these deficiencies might be supplied by further application of the time-honored method of comparing the clinical course of the disease with the changes found at autopsies, supplemented by bacteriologic studies made during life and confirmed after death.

With peribronchial pneumonia bronchi of medium size, on the cut surface of the lung, are surrounded by sharply defined zones of pneumonic consolidation perhaps 0.5 cm. in radius, and this lesion furnishes conclusive proof that the inflammatory process can extend directly through the bronchial wall reaching all alveoli within a limited distance for these alveoli bear no relation to the distribution of the terminal bronchi of the affected bronchus. This peribronchial pneumonia is usually characterized by fibrinous exudate, and pneumococcus has been found either in the blood of the heart or in the lung in all of 6 instances in which peribronchial consolidation has been recognized at autopsy; in half of these autopsies Pneumococcus Type II has been isolated and this relationship is especially noteworthy because Type II has been uncommonly associated with the pneumonias of influenza.

Lobar Pneumonia.—The distribution of lobar pneumonia has repeatedly furnished evidence that the process spreads like the peribronchial lesion directly through the tissue of the lung and is not necessarily disseminated by way of the bronchial tree. Pneumococci doubtless enter the lung by way of the bronchi; the occurrence of lobar pneumonia in frequent association with influenza which exhibits a peculiar capacity to destroy the defences of the lower respiratory passages is in harmony with this view. The presence of pneumococci in the blood furnishes no evidence that infection is hematogenous, for bacterial infections, particularly at their onset, are frequently accompanied by bacteriemia. The wave-like spread of lobar pneumonia may be indicated by a narrow zone of red hepatization separating a large patch of firm, gray consolidation from engorged but air containing lung tissue. A semicircular patch of consolidation not infrequently extends from the left lower lobe into the upper lobe at the site where the interlobular cleft is absent. This patch may be firm and gray in continuity with similar consolidation in the lower lobe but surrounded over its convex surface by a zone of red hepatization.

There is no reason to doubt that the lobar pneumonia which we have found with influenza has been constantly caused by pneumococci. We have encountered no instance of lobar pneumonia caused by the capsulated bacillus of Friedländer. The incidence of different types of pneumococci in the lung with lobar pneumonia has been as follows: Type IV, 32.4 per cent; Type II, atypical, 26.5 per cent; Type III, 17.6 per cent; Type II, 5.9 per cent; Type I, 2.9 per cent; no pneumococci found 14.7 per cent. It is noteworthy that this distribution of types is in sharp contrast with the lobar pneumonia of civil life with which Types I and II constitute the cause of two-thirds of all instances, and is in agreement with the etiology of the pneumonias found in an army camp (Funston) in the absence of influenza in epidemic proportion.

Bronchopneumonia.—Bronchopneumonia is associated with intense bronchitis penetrating to the finest bronchioles and is characterized by consolidation distributed in such definite relation to the bronchial tree that dissemination of the inflammatory irritant by way of the bronchi is evident. Consolidation occurs (a) in foci affecting alveoli in immediate proximity to the respiratory bronchioles and in consequence clustered about the terminal bronchi, the intervening alveolar tissue containing air; (b) in foci of the same character surrounded by intraalveolar hemorrhage which occupies all alveolar tissue between adjacent foci; (c) throughout whole lobules or groups of lobules, intervening lobules being unaffected; (d) surrounding bronchi of medium size like a sheath.

The lobar pneumonia of influenza is characterized by frequent association with purulent bronchitis and bronchopneumonia. The bronchopneumonia of influenza exhibits characters which serve to distinguish it from other forms of bronchopneumonia; (a) The associated lesions of the bronchi are unusually severe; purulent exudate accumulates within the lumen and the lining membrane is destroyed. (b) Pneumonia is frequently hemorrhagic with accumulation of blood within the alveoli and within and surrounding the bronchi. (c) There is unusual susceptibility of the injured bronchi and of the pulmonary tissue to secondary invasion by streptococci and staphylococci with consequent necrosis and suppuration. (d) Bronchiectasis frequently accompanies bronchitis. (e) Bronchopneumonia frequently fails to resolve and the lesion assumes the character of a chronic pneumonia.

With bronchopneumonia pneumococci are found with B. influenzæ in the bronchi and lungs in nearly half and in the blood in approximately one-third of instances of the disease, but hemolytic streptococci, staphylococci, S. viridans, B. coli, M. catarrhalis and other microorganisms are very frequently found in various combinations: they undoubtedly have a part in the production of the lesion. Mixed infection of the lung and even of the blood with pneumococci and hemolytic streptococci is often found, and study of the sputum during life has repeatedly shown that pneumococci alone are present shortly after the onset of the disease, whereas hemolytic streptococci appear later or are first discovered at autopsy. In such instances pneumococci have not infrequently disappeared from the lung and at autopsy hemolytic streptococci alone are demonstrable.

The part which B. influenzæ has in the production of bronchopneumonia is of great interest. This microorganism is demonstrable by cultures in at least three-fourths of all instances of bronchopneumonia but is obtained from the inflamed lung tissue in less than half. In no instance of pneumonia have we found B. influenzæ unassociated with other microorganisms, whereas repeatedly pneumococci have been the only microorganism demonstrable in the lung and very frequently the only organism present in the blood. In view of the difficulty of demonstrating the microorganism in plates overgrown by other bacteria, it is probable that its incidence in the bronchi is much higher, if it is not constantly present, whereas its isolation from the lung is in part referable to its presence in the small bronchi where it can be readily demonstrated by cultures or by microscopic preparations. We have been almost uniformly unsuccessful in demonstrating the microorganism in the alveoli of the lung. Goodpasture and Burnett,^[106] who have devised a special method for the demonstration of B. influenzæ in tissues, have found few of these microorganisms in the alveoli of the lungs.

Pneumonia characterized by the occurrence of small (peribronchiolar) spots of leucocytic pneumonia upon an almost homogeneous background of intraalveolar hemorrhage, was regarded by Pfeiffer as the characteristic lesion produced by his microorganism. B. influenzæ in our autopsies has borne the same relation to this lesion which it has exhibited to other forms of bronchopneumonia; pneumococci have been present with approximately the same frequency and hemolytic streptococci have often been found.

Streptococcus Pneumonia.—The occurrence of streptococcus pneumonia with suppuration occurring in the trail of influenza was frequently observed during the pandemic of 1889–90. It is now well recognized that the streptococcus concerned is one capable of causing hemolysis. Suppurative pneumonia referable to hemolytic streptococci is of two types which are readily separable by their anatomic characters: (a) One or several abscesses are situated below the pleura and accompanied by empyema. Their relation to severe lesions of the bronchi is not infrequently demonstrable, for a destructive lesion of the bronchial wall has penetrated into the surrounding alveolar tissue so that necrosis of tissue and subsequent abscess formation occur in continuity with the bronchial lumen. The localization of the abscess below the pleura is referable to the greater severity of the lesions of the small bronchi which are most numerous at the periphery, to the greater severity of these bronchial lesions at the bases of the lung, and to the relation of lymphatics within the interior of the lung to those of the pleura. It is not improbable that stasis of lymph caused by thrombosis of the lymphatics has a part in the production of abscess. Preceding or accompanying abscess formation, the lung tissue undergoes consolidation and in a wide area about the abscess has a homogeneous gray cloudy appearance occasionally mottled by opaque patches of necrosis. (b) Interstitial suppurative pneumonia is a lesion not infrequently found in association with influenza (21 times among 241 autopsies) and rarely, if ever, seen in its absence. There are few references to this lesion in the pathologic literature of the English language and those of German origin in great part refer to the period of the pandemic of 1889–90. The lesion is essentially suppurative lymphangitis, and both thrombosis and suppuration of the lymphatics are widespread throughout the affected lung. In proximity to the inflamed lymphatics and the surrounding interstitial septa, lung tissue throughout parts of the lobes or even throughout a whole (lower) lobe has undergone consolidation and has the gray, cloudy appearance of streptococcus pneumonia.

Staphylococcus Pneumonia.—Abscesses produced by staphylococci differ in anatomic characters and sequelæ from those caused by hemolytic streptococci. Small abscesses occur in one or several localized clusters; these abscesses are grouped about a bronchus and have their origin in its terminal branches. This relation may be readily demonstrated in microscopic sections. The lesion tends to remain localized and pneumonic consolidation is limited to the immediate neighborhood of the group of abscesses. There is no lymphangitis and the lesion is not accompanied by empyema.

Empyema.—Empyema is almost invariably associated with suppurative pneumonia caused by hemolytic streptococci. Among our autopsies purulent fluid has been found in the pleural cavity 55 times; it occurred 15 times among 178 instances of lobar or bronchopneumonia and 50 times among 60 instances of suppurative pneumonia referable to S. hemolyticus. In our experience hemolytic streptococci and pneumococci are the only microorganisms which exhibit a noteworthy capacity to penetrate from the lung to the pleural cavity. We have not found nonhemolytic streptococci (e. g., S. viridans) in association with empyema.

Staphylococcus has failed to invade the pleural cavity even when a pulmonary abscess has been present below the pleura, and in the only instances in which staphylococci have been isolated from the pleural cavity thoracotomy had been performed for empyema caused by hemolytic streptococci (2 instances) or an abscess communicating with both bronchus and pleura. B. influenzæ has been found in the pleural cavity with empyema only once and in this instance cannot be regarded as the cause of the lesion, for it has accompanied hemolytic streptococci.

Bronchiectasis.—Bronchiectasis has been frequently found as a sequela of the severe bronchitis of influenza and there has been abundant opportunity to study the lesion in process of development. These observations have furnished a satisfactory explanation of its etiology and pathogenesis. Infection of the bronchi by B. influenzæ, accompanied by a variety of other microorganisms, notably hemolytic streptococci and staphylococci, has caused profound changes in the bronchial wall beginning with destruction of the epithelial surface, and followed by necrosis penetrating partially or completely through the wall and occasionally extending into the surrounding alveolar tissue. The difference between the atmospheric pressure within the bronchi and the lower inspiratory pressure within the surrounding alveoli, accentuated by forced inspiration at intervals and by occlusion of the bronchioles with mucopurulent exudate, ruptures the necrotic tissue and produces longitudinal fissures which are recognizable both macroscopically and microscopically. In consequence of the separation of the edges of these fissures by intrabronchial pressure the circumference is increased. These rents in the wall are limited and partially healed by fibrinous pneumonia about them, by new formation of fibrous tissue from the bronchial wall, and adjacent interalveolar septa, by organization of fibrin within adjacent alveoli and finally by growth of epithelium over the denuded surfaces.

Bronchitis caused by B. influenzæ and pyogenic micrococci with necrosis of the bronchi wall is the essential factor in the production of bronchiectasis, but advanced bronchiectasis is found only in those individuals who have survived the onset of illness during several weeks, for dilatation under the influence of positive intrabronchial and negative extra-bronchial pressure occurs slowly.

Unresolved Bronchopneumonia.—Unresolved lobar pneumonia has not been recognized among instances of pneumonia following influenza, but unresolved bronchopneumonia is of frequent occurrence and has well definable gross and microscopic characters. There are purulent bronchitis, bronchiectasis and distention of the lung tissue, so that it fails to collapse; particularly characteristic are the indurated foci of peribronchiolar pneumonia, which being firm and sharply defined, have the appearance of miliary tubercles. When the process is sufficiently long continued there are recognizable patches of fibroid pneumonia. Microscopic examination shows that the lesion is characterized by organization of fibrinous exudate not only within the alveoli but within bronchioles as well, and by thickening of the alveolar walls, thickening of fibrous tissue about the bronchi and blood vessels, and thickening of interstitial septa. These changes may occur as peribronchiolar patches of consolidation, producing tubercle-like nodules, or may involve areas of hemorrhagic peribronchiolar or of lobular consolidation, or may be limited to the immediate neighborhood of bronchi (peribronchial).

No peculiarity of the bacterial flora of the bronchi or of the lung offers a satisfactory explanation of the failure of pneumonic exudate to resolve. Mixed infections have been common and S. hemolyticus, staphylococci, pneumococci, S. viridans, B. coli, etc., have been found in association with B. influenzæ but the incidence of these microorganisms has not been greater than with bronchitis. The lesion has occurred in association with B. influenzæ and pneumococci unassociated with other microorganisms. It seems probable that the severity of injury to the bronchial and alveolar walls accompanied by recurring bacterial invasion or by continued infection with B. influenzæ and one or several cocci, is the factor concerned in the inhibition of resolution and the production of chronic pneumonia. If the disease does not result in early death, chronic pneumonia has an opportunity to manifest itself.

In this investigation of the bacteriology and pathology of influenza and its complications, certain microorganisms have been found so frequently that it is desirable to discuss the pathogenicity of each and to define the character of the lesions which it causes.

Bacillus Influenzæ.—The microorganism has been constantly found in association with influenza when cultures and animal inoculations have been made from various parts of the respiratory tract within from one to five days after the onset of the disease at a time when there have been acute symptoms of the disease.

It is often identified with difficulty in the presence of other microorganisms and may be overlooked when a single culture is made. Repeated cultures from the throat alone made from the fourth to the eighth day after admission to the hospital, at a time when temperature had fallen to normal, have demonstrated the presence of B. influenzæ in 30.5 per cent, whereas the incidence of the microorganism in similar cultures on admission had been 63.4 per cent. The incidence of B. influenzæ in the present epidemic of influenza is not less than that found by Pfeiffer in the epidemic which he studied in 1892.

Nevertheless we have found that B. influenzæ is frequently an inhabitant of the mouth and throat of normal individuals. By inoculation of mice with the saliva or sputum of 76 patients with influenza, the microorganism has been found in 80.3 per cent; by inoculation of mice with the saliva of 185 normal men at army cantonments, it was found in 41.6 per cent; by inoculation of mice with saliva from 50 recruits immediately after they were assembled from isolated farming communities where only a few cases of influenza had occurred, it was found in 22 per cent. Figures for the same groups examined by a single throat culture were as follows: 65.7 per cent, 25.9 per cent and 0 per cent.

Experiments which we have performed on monkeys show that inoculation of the nasopharynx with B. influenzæ obtained from patients with influenza is followed by ill-defined symptoms associated with the presence of B. influenzæ within the throat. After from two to eleven days the symptoms and the microorganism disappear. Injection of B. influenzæ into the trachea causes bronchitis and the microorganism may be recovered from the inflamed bronchi two or three days after inoculation.

The constant association of B. influenzæ with influenza suggests that it is the cause of the disease. Its widespread occurrence in the throats of normal individuals does not contradict this view, since pneumococci long indistinguishable from those which usually cause lobar pneumonia are commonly found in the throats of healthy men. It is possible that B. influenzæ is a secondary invader, entering the respiratory tract when susceptibility is increased by an unknown virus causing influenza; but there is no convincing evidence in favor of this view. It is desirable to determine if microorganisms having the characters of B. influenzæ found with influenza differ in type from those found in the throats of healthy men and if the invasion of the respiratory tract by B. influenzæ is followed by the appearance of immunity reactions in the serum of the patient. Experiments on monkeys demonstrate the pathogenicity of the microorganism.

The relation of B. influenzæ to the bronchitis of influenza indicates that it has a part in the production of the pulmonary sequelæ of influenza. The microorganism has been found by a single culture from the bronchial passages in 80 per cent of instances of bronchitis with fatal pneumonia following influenza and is probably constantly present, usually in immense number, in the bronchial mucus. It is obtained from the pneumonic lung in only about 40 per cent of instances, and microscopic examination of prepared tissue shows that a bacillus with the morphology of B. influenzæ is often demonstrable in the bronchial passages but seldom in the alveoli of the lung. The microorganism is well adapted to multiply under conditions present in the bronchi but doubtless readily disappears from the alveoli which are the site of an inflammatory reaction. The microorganism has an important part in the production of the associated mucopurulent and hemorrhagic inflammation of the bronchi, but it is rarely if ever found in pure culture, being associated with a considerable variety of pyogenic cocci and occasionally bacilli. Infection of the bronchi with B. influenzæ in immense numbers offers an explanation of the severity of the inflammatory process within the bronchi, and of the subsequent dilatation and other chronic changes which occur in them. The presence of the microorganism and the accompanying injury to the ciliated epithelium and mucous glands are important factors in lowering the resistance of the bronchial passages to secondary bacterial infection.

We have obtained no evidence that B. influenzæ alone is capable of causing pneumonia. Its occurrence in less than half of all pneumonic lungs is explainable, in part at least, by its presence in the terminal bronchi which are cut across whenever the lung is punctured for culture. B. influenzæ alone has been found only once among 153 pneumonic lungs from which cultures were made, and in this instance (Autopsy 487) S. hemolyticus present in the blood of the heart, pleural cavity and bronchus doubtless had a part in the production of the associated pneumonia. Pfeiffer maintained that the lesion we have designated hemorrhagic peribronchiolar consolidation was characteristic of infection with his microorganisms. With this lesion B. influenzæ has been found in the lungs in slightly more than half of our autopsies but never alone, pneumococci being found in a third, hemolytic streptococci in more than a half and staphylococci in a fourth of the lungs examined.

B. influenzæ has relatively little capacity to penetrate from the bronchi into the lung tissue and rarely penetrates into the pleural cavity (once with Pneumococcus III, once with S. hemolyticus and once in pure culture), and only once has it been found in the blood of the heart, in this instance in company with S. hemolyticus. Capacity of the microorganism to penetrate from the bronchi into other tissues, both in man and as our experiments have shown in the monkey, is increased by association with pyogenic cocci.

Pneumococcus.—Lobar pneumonia following influenza, like lobar pneumonia in civil life unassociated with influenza, has been caused by pneumococci, but there is the notable difference that the pneumococci usually found are those types which are commonly present in the mouths of healthy men, namely, Types IV, III and atypical II and not the so-called fixed types, namely, Types I and II, which represent the usual cause of lobar pneumonia unassociated with influenza. It appears that influenza increases susceptibility to lobar pneumonia, so that it is frequently caused by microorganisms which under other conditions are less capable of producing this lesion. The association of the pneumococci usually found in the mouth with the lobar pneumonia of influenza does not exclude the possibility that pneumococci transmitted from one individual to another, when newly recruited troops are brought together, have an important part in the production of pneumonia.

Bronchopneumonia is frequently caused by pneumococci and the types which are recovered from the lung and blood do not differ from those found with lobar pneumonia, those usually present in the mouth being predominant, but the incidence of pneumococci with bronchopneumonia has been much less than with lobar pneumonia. Both lobar and bronchopneumonia caused by pneumococci have undergone secondary infection with hemolytic streptococci in a large proportion of instances and both pneumococci and streptococci are often recovered at autopsy. Nevertheless, the bacterial flora of the bronchi and lungs is much more varied with broncho than with lobar pneumonia, and it is evident that microorganisms other than pneumococci are capable of causing bronchopneumonia.

In instances of bronchopneumonia associated with pneumococci, fibrin has been abundant in the alveolar exudate.

The pneumococcus exhibits a notable tendency to produce an inflammatory process which extends through the bronchial walls and from one alveolus through the alveolar walls to those adjacent, for in 6 instances in which the bronchi were surrounded by pneumonic consolidation recognizable at autopsy, pneumococci were uniformly the causative agent, Pneumococcus Type II, otherwise rarely found, being present in half of these cases.

Pneumonia caused by one type of pneumococcus does not necessarily confer immunity from other types of pneumococci, and with somewhat limited opportunity we have observed a number of instances in which, following recovery from pneumonia caused by one type of pneumococcus, a second attack of pneumonia, usually fatal, has been associated with pneumococci of a different type. This recurring pneumonia in a considerable proportion of the relatively small number of instances observed has been produced by Pneumococcus Type II which otherwise has been seldom found among the cases which we have studied. The virulence of this microorganism doubtless explains its ability to cause recurrent pneumonia.

Streptococcus Hemolyticus.—Secondary infection with S. hemolyticus is a common event during the course of lobar pneumonia following influenza. It is noteworthy that this streptococcus infection of the lung has almost invariably occurred in the stage of red hepatization, whereas with gray hepatization, when the alveoli are filled with polynuclear leucocytes, S. hemolyticus rarely invades the lung. It is possible that infection with S. hemolyticus tends to prolong the stage of red hepatization.

The most significant change produced in the pneumonic lung by streptococci is necrosis. When after death with lobar pneumonia hemolytic streptococci, usually associated with pneumococci, are found both in the lungs and blood of the heart, the lung contains patches of necrosis recognized microscopically, in which the alveolar walls and exuded cells have uniformly lost their nuclei. Microscopic examination demonstrates the presence of chains of streptococci in immense number in these necrotic foci; elsewhere chains of streptococci occur but are much less abundant. In some instances streptococci exhibit a tendency to enter lymphatics and to cause acute lymphangitis with lymphatic thrombosis and edema of the adjacent interstitial tissue.

Hemolytic streptococci have been more frequently found in association with broncho- than with lobar pneumonia. In 24.5 per cent of instances of lobar pneumonia, doubtless in all instances caused by pneumococci, hemolytic streptococci have invaded the lungs and in 12.6 per cent of instances have found their way into the blood. With bronchopneumonia hemolytic streptococci have been obtained from the lungs in 29.8 per cent of instances and from the blood of the heart in 34.3 per cent.

With lobar pneumonia there is little doubt that pneumococcus has been the primary cause of pneumonia, but with bronchopneumonia pneumococci have been less frequently found. It is difficult to determine how often hemolytic streptococci have invaded a bronchopneumonic lesion, caused by pneumococci because pneumococci tend to disappear. In numerous instances in which the sputum had been studied during life, it was evident that pneumonia was primarily referable to pneumococci, and hemolytic streptococci made their appearance in the sputum late in the disease or were first recognized at autopsy.

When hemolytic streptococci occur in association with bronchopneumonia, foci of pulmonary necrosis similar to those found under the same conditions with lobar pneumonia have been repeatedly found by microscopic examination. In the patches of necrosis, cocci in chains are much more abundant than in the tissue elsewhere.

In some instances of pneumonia, caused by hemolytic streptococci, opaque gray or yellowish gray patches of necrosis occur upon a background of flaccid homogeneous consolidation which has a peculiar cloudy, gray color. This mottled consolidation may implicate an entire lower lobe and has the characteristic features neither of lobar nor of bronchopneumonia. More frequently the lesion is less widespread and necrosis occurs in one or several spots which undergo softening so that finally a small abscess cavity may be formed; it is surrounded by pneumonic consolidation which is soft and has the cloudy appearance described above. These pulmonary abscesses are almost invariably situated below the pleural surface; the adjacent pleural cavity is infected by streptococci and there is purulent inflammation of the pleura.

Streptococcus infection, which has been described, doubtless has its origin in the bronchi, for in favorable sections it is not infrequently possible to demonstrate that necrosis extends through the bronchial walls into the surrounding alveolar tissue and is followed by suppuration with abscess formation. Localization of abscesses below the pleura is in part at least referable to transmission of streptococci by way of the lymphatics.

Streptococci in the lung, as in other tissues, often invade lymphatics and produce an acute inflammatory reaction within and about these vessels. The peculiar lesion which may be designated suppurative interstitial pneumonia is a suppurative lymphangitis associated with inflammation and edema of the interstitial tissue. Lymphatics invaded by streptococci are the site of acute lymphangitis; occlusion by fibrinous thrombi occurs and finally the immensely distended lymphatics, filled with purulent fluid, take a characteristic nodular or beaded form and pus flows from them when they are cut. Streptococci are present in vast numbers. Suppurative inflammation may extend to the surrounding interstitial tissue which is distended by inflammatory edema. This interstitial suppurative pneumonia extends up to the pleural surface and empyema is almost invariably associated with it. The lesion is seldom seen in the absence of influenza.

One of the most significant characters of S. hemolyticus is its ability not only to enter the bronchi and penetrate into the tissue of the lung, but to find its way into more distant structures, namely, the pleural cavity, pericardial sac and peritoneal cavity and to penetrate into the blood. Among 121 examinations, hemolytic streptococci were found in the bronchi in 47.9 per cent; among 153 examinations of the lung it was present in approximately the same proportion, namely, 50.3 per cent; among 218 examinations of the blood it was found in 39 per cent. In 4 of 5 fatal pneumonias in which the organism has penetrated into the bronchi it has ultimately found its way into the blood.

Nonhemolytic Streptococci.—In contrast with S. hemolyticus nonhemolytic types have rarely been encountered in association with the pneumonias of influenza. S. viridans has been found only 5 times among 153 autopsies in which cultures have been made from the lung and has been invariably associated with other microorganisms. In no instances have nonhemolytic streptococci been found with empyema. In one autopsy with lobular bronchopneumonia S. viridans has been isolated from the blood of the heart and in this instance it has been found in the bronchus and lung as well. This type of streptococcus is evidently little adapted to invade the bronchi and produce lesions of the lung and adjacent tissues.

Staphylococci.—Staphylococci have been very frequently isolated from the bronchi in association with the pneumonias of influenza, being found in approximately half of our autopsies. Their isolation in cultures from the lung in a fourth of the autopsies examined is in part perhaps referable to their presence in the small bronchi cut across when the lung is punctured for cultures. S. aureus shows little ability to invade the pleura, being found in association with empyema only 3 times; in these autopsies there has been opportunity for entrance from the exterior through thoracotomy wounds in 2 instances and from a bronchus in free communication with an abscess which had ruptured into the pleural cavity in 1 instance.

Abscesses of the lung caused by staphylococci have been found in a small number of autopsies and have exhibited characters which differ from those ordinarily seen in association with S. hemolyticus. Small, sharply defined abscesses are grouped about terminal bronchi, so that they occur in one or several isolated clusters. Microscopic examination demonstrates that these abscesses have arisen by destruction of the bronchial walls and extension of suppuration into the surrounding alveolar tissue; clumps of staphylococci are found in sections through the abscess, and cultures made from the pus within the abscess cavity demonstrate the presence of S. aureus or albus, but the microorganism may be missed if the culture is made from the adjacent lung tissue. It is noteworthy that there is little tendency for the staphylococcus to infect the pleura for even though these clusters of abscesses have been situated just below the pleura, there has been no associated empyema.

Staphylococci have scant tendency to enter the blood and have been obtained from the blood of the heart only once, in this instance with hemolytic streptococci.

Pneumonia of Measles.—Pneumonia following measles has been responsible for a considerable part of the deaths occurring in the United States Army during the period of the war. The importance of measles as a factor in the production of pneumonia is illustrated by the history of pneumonia at Camp Funston from the establishment of the camp in September, 1917, until September, 1918. Pneumonia following measles occurred throughout the year; but in association with the high incidence of measles during the second half of November and the first half of December, 1917, there was an outbreak of related pneumonia characterized by frequent empyema and a mortality of 45.3 per cent.

During the period of our investigation at Camp Funston there were 112 cases of measles, but no pneumonia occurred among them. At Camp Pike, during the period of observation, there was an outbreak of measles almost coincident with the epidemic of influenza, and among 867 cases pneumonia occurred in 56, otitis media in 48, and mastoiditis in 23. Pneumonia following measles was almost coincident with that of influenza, and it is not improbable that the epidemic of influenza had an important part in the production of pneumonia in individuals suffering with measles.

In 9 of 56 instances of pneumonia following measles at Camp Pike, S. hemolyticus had invaded the lung and caused pneumonia; among 48 instances of otitis media following measles a very large proportion were caused by hemolytic streptococci, and 21 of 23 instances of mastoiditis were caused by the same microorganism. No complication caused by S. hemolyticus occurred among 37 patients who carried this microorganism when admitted to the hospital.

A special study has been made to determine if those patients with measles who carry S. hemolyticus in their throats are especially susceptible to complications during the course of measles. The low incidence of streptococcus “carriers” among those admitted to the hospitals with measles was noteworthy both at Camp Funston (2.67 per cent) and at Camp Pike (4.2 per cent). Indeed, it was found at both places that the incidence of hemolytic streptococci in the throats of normal men in the camp was higher (Camp Funston 21.9 per cent; Camp Pike 7.4 per cent) than that in the throats of those admitted with measles. While in the hospital there was a gradual increase of the incidence of S. hemolyticus, so that in three weeks it had risen to 19 per cent at Camp Funston and to 26.2 per cent at Camp Pike. It seems not improbable that hemolytic streptococci disappear from the throat in the early stages of measles, so that they are not demonstrable by cultural methods. During the course of the disease in the hospital ward the number of those with S. hemolyticus has increased in some wards with great rapidity, infection being apparently transmitted from one individual to those adjacent. At Camp Funston the incidence of S. hemolyticus in the throats of those convalescent with measles was almost identical with that among normal men in organizations from which the patients had come, but at Camp Funston the percentage of hemolytic “carriers” among convalescents was much higher than that obtained among normal men in the camp.

The demonstration of S. hemolyticus in the throat of a patient suffering with pneumonia is not conclusive proof that the lungs have been invaded by this microorganism. Pneumonia in individuals carrying S. hemolyticus in the throat may pursue a favorable course and exhibit no evidence that the microorganism has found its way into the lung. In some instances hemolytic streptococci have been found in the bronchi at autopsy yet none have entered the lung or blood and the lung exhibits none of the lesions which are referable to hemolytic streptococci. Nevertheless, the occurrence of S. hemolyticus in cultures from the throat of a patient with pneumonia suggests the probability that he is suffering with streptococcus pneumonia.

Pneumonia following measles studied in 18 autopsies upon patients who died during or shortly after the epidemic of influenza, exhibited all the characters exhibited by the pneumonias of influenza. In 4 instances there was typical lobar pneumonia; bronchopneumonia was found in all but 3 instances, being associated with lobar pneumonia twice. All the noteworthy features of the bronchopneumonia of influenza have been reproduced among these instances of pneumonia with measles; there is severe injury to the bronchi, and purulent bronchitis has been present in 13 instances; pneumonia has frequently had a hemorrhagic character, hemorrhagic peribronchiolar pneumonia occurring in 5 instances; secondary infection of the pneumonic lungs with hemolytic streptococci has been common; bronchiectasis has been associated with bronchitis (in 8 instances) when purulent bronchitis has persisted several weeks; and unresolved bronchopneumonia has been more frequent (6 instances or one-third of the autopsies) than with influenza.

The bacteriology of pneumonia following measles has been the same as that of influenzal pneumonia. B. influenzæ is found with few exceptions in the bronchi and much less frequently in the pneumonic lungs.

Pneumococci have been obtained from the blood or lungs in 5 of 13 instances of lobar or bronchopneumonia unaccompanied by suppuration; when suppuration has been absent no hemolytic streptococci have been found. Pneumococci concerned in the production of pneumonia of measles, as with influenzal pneumonia, have been types usually found in the mouth; Pneumococcus II atypical has been found 6 times, Type IV once, Type I once.

Hemolytic streptococci have invaded the pneumonic lung in 5 instances. They have produced subpleural abscesses accompanied by empyema in 2 instances. Interstitial suppurative pneumonia, a lesion repeatedly found in consequence of secondary infection with S. hemolyticus following influenza and rarely found in this country, at least in the absence of an epidemic of influenza, has occurred 3 times among 18 instances of pneumonia following measles.

The foregoing observations show that the pneumonia following measles, which has occurred almost coincidentally with pneumonia accompanying epidemic influenza has reproduced the lesions found with influenzal pneumonia. They indicate that influenza attacking patients with measles has had a part in the production of this pneumonia.

The Transmission of Streptococcus Pneumonia.—The importance of streptococcus as a cause of pneumonia following influenza was recognized during the pandemic of 1889–90. Patients suffering with pneumonia following influenza or measles are susceptible to infection by S. hemolyticus and this streptococcus pneumonia may be transmitted from one patient to another throughout a ward in which patients with pneumonia are assembled. There is no evidence that primary pneumonia caused by S. hemolyticus has prevailed as an epidemic in the army or elsewhere in the absence of preceding infection with influenza or measles.

Our autopsies demonstrate that at least half of all deaths which have occurred at Camp Pike have been caused by hemolytic streptococci which have invaded the lung and entered the blood. It is significant that this mortality had its origin in the first half of the epidemic of influenza at a time when the military and medical organization of the camp was confronted with an unforeseen emergency which overwhelmed all agencies for the care of disease. Curves prepared by referring cases of pneumonia in which autopsy demonstrated the nature of the fatal infection back to the date of the onset of influenza, demonstrate that fatal streptococcus pneumonia was frequently acquired during the early period of the epidemic, the maximum number of cases occurring September 23 and 24 and became gradually less common as a sequela of the influenza which began at a later period. Fatal pneumococcus pneumonia had its origin with increasing frequency at a later period, the maximum incidence following influenza which had its onset September 29 and 30. Overcrowding of influenza patients in infirmaries, ambulances and hospital had an important part in the dissemination of streptococcus pneumonia among influenza patients whose disease might otherwise have pursued a benign course.

The most important factor in the high incidence of streptococcus pneumonia has been the spread of the disease in the hospital wards. On September 24 the base hospital contained 2,789 patients, although it had been planned to care for only 2,009. With the progress of the epidemic the number of admissions increased very rapidly, so that on September 30 the hospital contained 3,587 patients and on October 5, 4,233. After September 24 the milder cases of influenza were treated in barracks. The pressing need of diminishing the overcrowding of the hospital was fully recognized and adjacent barracks were transformed into hospital wards; between October 3 and 6, 1,362 patients were transferred from the hospital to these quarters.

In the main hospital, during the period of overcrowding 20 wards for patients with pneumonia were added to the two which already existed. These hastily organized and overcrowded wards have been attacked by outbreaks of streptococcus pneumonia, which during certain periods have been fatal to more than two-thirds of those who have been admitted with pneumonia, whereas in the two long established wards for pneumonia isolated cases of streptococcus infection, which have appeared, have failed to spread to other patients and pneumococcus pneumonia with few exceptions has been found in those who have died. In one newly established ward 67.5 per cent of those admitted within a period of three days have died, and in all of the 23 autopsies which have been performed, streptococcus pneumonia has been found. In another ward 50 per cent of all who have been admitted during a period of one week have died, and among the autopsies performed on these individuals pneumococcus pneumonia has been found in 6 and streptococcus pneumonia in 14. The sputum of 9 patients in this ward has been examined on admission, and pneumococci, but no streptococci, have been found. All these patients have died, and infection with S. hemolyticus has been found at autopsy in 7.

Transmission of Pneumococcus Pneumonia.—Our study of secondary ward infection has not only shown that patients with pneumococcus pneumonia following influenza are susceptible to infection by S. hemolyticus, but that patients suffering with pneumonia caused by one type of pneumococcus may be infected with another type during the course of the disease or after convalescence has begun, the second infection being acquired from patients in adjacent beds. Pneumonia caused by Type IV has ended in crisis and has been followed by a period of normal temperature; recurrent pneumonia has been fatal and Pneumococcus Type II has been found in the organs at autopsy. Pneumonia caused by Type I has been followed by recurrent pneumonia caused by Pneumococcus II atypical acquired from a patient in the next bed. These secondary pneumococcus infections acquired within the hospital are apparently not uncommon.

Prevention of the Transmission of Pneumonia.—The essential factor in the management of influenza and pneumonia is such isolation of each patient that microorganisms cannot be transmitted from one to another or from attendants or others to patients. This condition may be fulfilled by the separation of patients in rooms or isolated compartments especially constructed for the treatment of pneumonia and by the employment of all possible means to prevent the transmission of infection from one patient to another by physicians, nurses and orderlies. It is desirable to examine attendants to determine if they carry hemolytic streptococci in their mouths and to exclude those who are found to be “carriers.”

Influenza is a self-limited disease which, in the absence of complications implicating the lower respiratory tract, is of relatively mild character. When death occurs as the result of influenza it is with very rare, if any, exceptions referable to pneumonia; we have invariably found pneumonia in those who have died in consequence of influenza. The individual attacked by influenza may carry within his upper respiratory passages pneumococci or hemolytic streptococci capable of invading the bronchi and causing pneumonia, but in most instances the microorganism which produces serious pulmonary complications is derived from others with whom the influenza patient has come into contact. The greatest source of danger to one with influenza is contact with patients who have acquired pneumonia, and this danger is immensely increased when infection with S. hemolyticus makes its appearance among pneumonic patients. Hospital epidemics of streptococcus pneumonia will be prevented when the disease is dreaded as much as puerperal fever or the hospital gangrene of former years, and widespread knowledge of the suppurative pneumonias of influenza will bring a clear recognition of the fatal character of streptococcus infection in patients suffering with pneumococcus pneumonia.

Overcrowding of barracks has been an important factor in the propagation of acute respiratory disease and in the transformation of otherwise trivial influenza into fatal pneumonia. Crowded troop trains have doubtless had a part in disseminating infection among newly assembled recruits. Should these dangers be recognized they may be avoided by appropriate measures which will promote rather than retard those military aims which must be placed foremost in time of war. It may be possible by adequate expenditure to avoid the death of thousands of recruits within one month of their entrance into military service.

A second factor in the increase of death rate from pneumonia is the overcrowding and confusion of hospital facilities in the presence of an epidemic disease. When troops are maintained in camps precautions should be taken to provide effective safeguards against the overcrowding of the base hospital.

Isolation of each patient with pneumonia is the most effective way of protecting him from infection and of preventing him from becoming a possible source of danger to others. The effectiveness of this isolation will depend upon the separation of patients by some means more effective than the cubicles composed of sheets heretofore employed, upon an aseptic technic sufficiently rigid to prevent the transfer of pyogenic infection to pneumonia patients, and upon the exclusion from the ward of those who harbor S. hemolyticus.

Even should each patient be completely isolated from his neighbors, no effort should be neglected to determine, as far as possible, the nature of the infection with which he suffers. In the presence of an overwhelming epidemic such as that which attacked our army camps, the bacteriologic work which is required may be far beyond the facilities which are available and in many instances it may be wholly impossible. Nevertheless effective control of streptococcus pneumonia will depend upon its recognition as soon as it appears, and bacteriologic examination of the sputum offers the readiest means for its identification. The routine performance of autopsies will furnish an index of the success of the measures in force, and the discovery of suppurative pneumonia will suggest the presence of imminent danger.

However perfect the organization of pneumonia wards and however accurate the aseptic technic in force, it is desirable to separate as far as possible those infected with streptococcus from those who are free from this infection, so that the accuracy of the technic in force may not be put to too severe a test. When streptococcus pneumonia has appeared in a ward it should be closed to further admissions.

Those who are concerned in the planning and construction of military and other similar hospitals might well give special attention to the possibility of epidemics such as those which we have experienced, and special provision might be made to avoid overcrowding in the presence of a demand far in excess of the routine need for hospital facilities. In the construction of these hospitals appropriate provision should be made for the care of patients with pneumonia. Medical officers should receive detailed instruction in the organization and conduct of wards designed for the treatment of pneumonia.