U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter III




In order to portray the pathology of the respiratory diseases of the World War it is necessary to classify the lesions and to define the meaning of certain descriptive terms so that misunderstanding may be minimized. While it is impossible in the present state of our knowledge to separate the lesions according to the organisms producing them, certain conditions found in the respiratory tract appear to have been associated with some organisms more frequently than with others. Bacteria, therefore, must be considered, and it is convenient to describe a type of pathological change as most characteristically produced, in so far as careful and painstaking bacteriological research of the period of this war has divulged, by one organism. For instance, interstitial pneumonia, as described by MacCullum,1 is the result of infection of the lung with a hemolytic streptococcus, although there were found numerous changes of interstitial type in lungs from which this organism was not isolated by methods which ordinarily showed its presence, other organisms, as Streptococcus viridans or Group IV pneumococci, apparently acting as the etiologic agents.

By force of circumstances we are unable to portray the acute respiratory lesions which did not prove fatal. While undoubtedly persons in the early stages of acute respiratory disease died from other causes, no observations have been found on the pathology of the respiratory tract in such cases. Certain assumptions are based on the condition of tissues at death which seem to indicate the sequence of events leading up to the final condition. Furthermore, it is manifest that the complete sequence can be surmised only from the study of lesions observed at varying periods after the onset of clinical symptoms. Here variations in judgment are bound to occur as the variables are numerous, including, as these do, the virulence of the organism, its numbers, the resistance of the host both physically and chemically, and those accidents which, on the one hand, favor the patient, on the other, assure a fatal outcome of the disease. Examples are the penetration of a pulmonary vein by an abscess and superimposed infections with additional species of organisms. The previous experience of those who made the examinations and of those who interpreted them is also responsible for variations not only by reason of the amount of this experience but also its character. The endeavor is therefore made to delineate a picture which will be clear and interpretable in the light of both present and future knowledge.

The specimens which form the basis of the illustrations were selected from the collections of the Army Medical Museum and were obtained in the necropsies of over 600 cases at various Army stations during and immediately following the World War. The fatal respiratory lesions were considered primary by those writing the protocols in a relatively small proportion of the cases. The majority were considered secondary to clinical influenza, measles, or other acute respiratory infections, either not definitely diagnosed or diagnosed as acute pharyngitis,


bronchitis, tonsillitis, or rhinitis. The indefinite diagnoses, acute pharangitis, bronchitis, tonsillitis, and rhinitis were found rarely in the material sent to the Army Medical Museum, as most of the specimens were collected during waves of acute respiratory disease of epidemic proportions.

The pathology of measles independent of its serious complications will not be described, as the material studied offers no possibility of clearly differentiating it from the secondary infections responsible, in large part at least, for death. A few observers were of the opinion that the virus or etiologic organism causing measles produced a characteristic picture in the lungs.2 A similar picture was described as probably due to the influenza bacillus and indeed a similar organism was present in cases complicating or following measles. Since, clinically, both measles and influenza were prevalent at the same time it is quite possible that patients with measles were affected also by influenza. It is impossible, in the light of present knowledge, to determine where the lesions due to the cause of influenza left off and those of secondary invaders began, but in certain instances the sequence of events leading to death occurred so rapidly as to make the process clinically and anatomically a disease entity. It is quite possible, however, that the secondary invader produced its lesions synchronously with those of the primary, the latter acting either to increase the virulence of the secondary or to depress the resistance of the host.

In the larger waves of acute respiratory disease in the spring and fall of 1918, and in the fall of 1917, an acute infection with measles or influenza-like disease was followed by varying numbers of some type of pneumonia. In a few of the measles cases pneumonia occurred after discharge from hospital, but in most instances this disease was manifest during hospitalization and while measles was present. The onset of the pneumonia was sometimes simultaneous with that of measles but more often followed the drop in temperature at the stage of full eruption.

In the cases diagnosed clinically as influenza, there was also a lack of uniformity in the time of onset of pneumonic symptons. The fulminant cases appeared to have serious pulmonary involvement from the start, while others, though showing the prostration and other symptoms of the clinical disease, did not have definite pulmonary signs for a varying time afterwards.

Cases which were fulminant from the start were the ones in which culture showed B. influenzae most often, and in which this organism was found most frequently alone in the culture from the lungs. Such cases occurred, according to available data, rarely in the fall of 1917, more frequently in the spring epidemies of 1918, and most often during the pandemic of the early fall of 1918. The same anatomical picture varying in the extent of the process in the lungs and in the stage of the lesion was found in some part of the pulmonary parenchyma in most of the lungs of the pandemic wave and in many of those from the spring epidemic. The material available is too scanty to enable one to judge of its frequency in the fall of 1917, but it did occur. Lesions in which other organisms, pneumococci, streptococci and staphylococci appeared to be responsible varied in time of appearance of clinical symptoms, in extent of tissue involved, and in severity. Some of them (pneumococcus Type II at Camp Grant) invaded early in the course of the pandemic infection and some fulminant


cases appeared to be attributable to these organisms alone, though this was the exception.

The lesions in the lung at necropsies, whether such cases were recorded as primary or secondary, were characterized by their extreme variety not only in different cases but also in the same case. It was frequently difficult to decide which type of pneumonic process predominated in the lungs of a single case, and the more carefully these organs were sectioned, the more difficult the decision. For example, a lower lobe might show a large part uniformly consoli- dated by a fibrinous exudate of the character usually seen in lobar pneumonia, the peripheral portions lax and studded with small peribronchiolar consolidations, in the lower lobe on the other side a pneumonic consolidation surrounding and spreading out from the bronchus nearest the vertebrae, in the balance a scattering or confluent lobular consolidation, while the upper lobes contained scattered lobular and peribronchiolar consolidations and strie of interstitial lymphangitis about bronchi and in interlobular septa.

The age of the lesions as judged by the gross and microscopic appearances, also varied and often to extreme degrees, so that it was evident to the pathologists that in most of the fatal cases at least there had been several periods of advance, or the infection, originating in one part of the lungs, had spread to another and then another part, or else other organisms than those causing the initial process had gained entrance and, in turn, had produced pulmonary inflammation. The organisms recovered were often of different species and varieties, which accounts in large part for the varied bacteriological findings at necropsy and probably also the changes which occurred in the bacterial flora as determined by culture of the discharges from the respiratory tract during life. In fact, the varied bacterial flora, with the differences in the media required for their growth, made it extremely difficult to ascertain the bacterial cause or causes of the pneumonic lesions. The epidemic proportions in which these diseases appeared made it physically impossible for the laboratories to carry out sufficiently extensive cultural work to determine the organisms concerned with any degree of certainty, though at most camps the predominant ones in the etiology of the secondary pneumonia undoubtedly were recognized.

It was difficult in many instances to determine the boundaries of the pathological anatomical processes, yet certain distinct types or distributions did occur as the only ones in a given lobe or lung, or were so separated from others as to render differentiation possible and to enable the pathologist to determine the organisms they contained as shown by smear, culture, and microscopical section of the tissue. In the illustrated description which follows, the lesions are considered according to anatomical type and are preceded by definitions descriptive of these types.


Typical lobar pneumonia was present in all camps and was particularly noted during the fall and winter of 1917-18. Its exact incidence can not be determined, since the clinical diagnosis, lobar pneumonia, frequently was proven incorrect at necropsy. The morbidity, as judged by a study of the reports in the literature from the Army camps and the protocols at the Army Medical


Museum, was not high, and the case fatality rate was low. The condition followed measles and acute epidemic respiratory disease but was not nearly as frequent as other types, and if found, usually was associated with other varieties of pneumonia in lobes of the lung not affected by the lobar process. In order to separate the anatomical types of pneumonia one from the other, a rather narrow interpretation has been made of the anatomical characteristics used as

FIG. 1.- Postinfluenzal pneumonia. Typical lobar pneumonia in the right upper lobe at the stage of gray hepatization. Confluent bronchopneumonia in the right lower lobe surrounding two of the larger bronchi. In the vertebral portion of the right lower lobe are seen numerous peribronchiolar foci of consolidation with confluent lobular pneumonia. Acute hemorrhagic bronchitis of both lungs, congestion of the left upper lobe, peribronchial consolidations in the left lower lobe. Just beneath the bifurcation of the trachea there is an abscess involving the lymph nodes of the mediastinum. Pneumococcus Group IV was isolated from the sputum and lung. Sections of tissue show streptococci in the abscess of the interlobular septum in the right upper lobe and Gram-negative bacilli in the peribronchiolar consolidations. Accession number 2711, Army Medical Museum. Negative number 30583

criteria of the diagnosis, lobar pneumonia. The pathologic anatomy of the condition has been described in detail so that it may be easily differentiated from the other pneumonia types.

Lobar pneumonia is an inflammation of the lung characterized by the uniform consolidation of a whole lobe or practically all of a lobe. (Fig.1.) Three stages are presented in this type.


In the first, or congestive stage, the lung is heavy but compressible, there is little or no crepitation, it is moist, red in color, drips blood on section and the markings are visible. It is not firmly consolidated but collapses slightly if at

FIG. 2.– Dilated small vessel in the congestive stage of lobar pneumonia. Capillary vessels are engorged and there is an escape of red blood corpuscles, serum, a small amount of fibrin, and fairly numerous leucocytes into the alveoli adjudacent to the small vessel. The lymphatics surrounding the vessel are distended with leucocytes, which infiltrate the alveolar walls. Accession number 22728. Army Medical Museum. Negative number 41224. Hematoxylin and eosin stain; X 400

all. The vessels are engorged with blood which distends the alveolar capillaries, increasing the thicknesss of the alveolar walls. The alveolar walls are also thickened by edema, the epithelium is more or less desquaniated, and the


alveoli contain serum, red blood corpuscles in varying numbers, and a few leucocytes. Strands of fibrin form early but do not dominate the picture at this stage. (Fig. 2).

In the second or red hepatization stage, the lung is firmly consolidated. It fills the pleural cavity and usually shows the rib markings. It is friable and dark purplish red in color. On section the surface is dry, a dull dark-red in color. Plugs of firm exudate can be expressed with the knife edge. The markings are obscured but bronchi of medium and large size are distinct and do not project above the surface. The degree of completeness of the consolidation is common to practically all of the tissue involved though the process is usually slightly more advanced about the larger bronchial and vascular trunks near the hilus. The larger vessels are filled with blood but the smaller vessels and alveolar capillaries are compressed to a greater or lesser degree by the exudate which now fills the alveoli, compressing the walls. The exudate usually contains many leucocytes and a dense fibrin mesh has formed, producing a clot of the entire exudate mass. Red corpuscles retain their stain and vessels still contain an excess of blood. (Figs. 3 and 4.)

In the third or gray hepatization stage, the lobe affected is large, does not collapse, is firm and usually shows rib markings. The color is lighter than in the second stage and may be yellowish gray. It is friable to a greater extent usually than in the preceding stage. On section the cut surface is plane, dull and dry. It is usually somewhat mottled a reddish-gray to yellowish-gray. Firm plugs may be expressed by the knife edge. Because of the lighter color the bronchial walls do not stand out distinctly. The process is uniform for practically all of the tissue affected except that it is slightly more advanced near the hilus where, in the lungs typical of this stage, some liquefaction of the exudate may have occurred. The alveoli are distended with exudate composed of a dense mass of leucocytes in a meshwork of fibrin. Red blood corpuscles do not stain well and the alveolar walls are compressed to such a degree that it is difficult to see them or to make out the capillaries. In the later stages the exudate becomes necrotic and autolyzed as solution takes place and this is accompanied by a filling of the capillaries with blood. (Figs. 5 and 6.) This stage passes over gradually into that of resolution which is somewhat less uniform as the resumption of the full blood and lymphatic circulation is irregular. Lesions corresponding to the description of the three stages but not involving practically all of a lobe are not designated as lobar. Neither are lesions of the same type in lobules which show varying stages in the process even when the greater part of a lobe was thus consolidated.


Bronchopneumonia is a pneumonic consolidation spreading outward from the bronchi a varying distance into the surrounding alveoli, the oldest process being in the vicinity of the bronchus situated in the approximate center of the lesion. Macroscopically, a lung so affected is irregularly consolidated; usually both lungs are affected. On section, the consolidations surround the bronchi extending irregularly out into the parenchyma. When two or more neighboring or adjacent bronchi are affected the lesions may coalesce to form


one large pneumonic mass, the lines of division into separate components being difficult to determine macroscopically, though microscopically variations in the

FIG. 3.- Alveolus in the early stage of red hepatization, showing capillary engorgement, nearly complete desquarnation of the epithelium, and an alveolar exudate in which fibrin is abundant. Swollen epithelial cells are seen along the alveoiar wall and free in the serum-filled space between the fibrinous mass and the alveolar wall. Accession number 2272, Army-Medical Museum. Negative number 41217. Hematoxylin and eosin stain; X 600

stage of the process in different areas can usually be made out. The character of lesion varies widely and may show the same stages as the lobar type but is more irregular, the oldest lesions being near the bronchi. Hemorrhagic areas


are frequent and may predominate or constitute practically the entire gross picture. As a rule, there is less fibrin formation and the distribution is irregular 

FIG. 4- Stage of late red hepatization stained for retictilurn and show ing the dilated capillaries in the alveolar walls. Accession number 3115, Army Medieal Museum Negative number 45556. Reticulum stain; X250

Sections of small areas frequently show a wide variation in the character of the exudate between neighboring alveoli. (Fig. 7.)

There are two types of pneumonia which answer the description above. In one the bronchi are the site of a severe inflammation usually purulent in


character. The epithelium is largely desquamated, the wall densely infiltrated with exudate, often predominantly purulent and sometimes the wall is necrotic. The process appears to have extended through the bronchial wall to the surrounding alveoli. In the second, the bronchial mucosa is little affected in the earlier stages. The peribronchial and perivascular tissues are densely infil-

FIG. 5- Alveolis of early gray hepatization. The exudate as contracted as the result of fixation. Alveolar walls are compressed some capillaries still being filled with red blood corpuscles. Some of the epitheliurn is still adherent to the alveolar wall. Accession number 22728, Army Medical Museum. Negative number,41223. Hematoxylin and eosin stain: X 400

trated with exudate, usually purulent, but, the leucocytes are often predominately lymphocytic and the inflammation appeats to have extended primarily as, a lymphangitis along the peribronchial and perivascular lymphatics and then extended out into the surrounding alveoli. In the later stages the inflammation extends through the bronchial wall to the mucosa which is thus secondarily affected.



Groups of lobules, occasionally single isolated lobules, are affected. Approximation of groups may cause most of a lobe to be involved. Grossly there

FIG. 6.- Lung in gray hepatization, showing compression of the alveolar walls and capillaries. Accession number 3115, Army Medical Museuim. Negative number 45562. Reticulum stain; X 250

are scattered nodular areas of varying size throughout one, usually many lobes. Where these nodules reach the pleural surface as in those of the periphery of the pulmonary parenchyma, the outlines of the lobtules clan be muade out as a


mosaic-like pattern on the pleural surface. The individual lobule is usually uniformly consolidated though the lesions microscopically appear more advanced near the bronchiole of the lobule. The nodules vary in firmness and in

FIG. 7.- Postinfluenzal pneumonia. Spreading bronchopneumonia of left upper lobe and upper portion of left lower lobe. Peribronchiolar foci in base ofleft lower lobe. Confluent lobular pneumonia of right upper and middle lobes. Peribronchiolar foci in right lower lobe. Pneumococcus,Group IV, in sputum. Streptococci and bacillus of Friedlander in cultures from the bronchi. Accession number 1433. Army Medical Museum Negative number 30649

friability. Hepatization as seen in the lobar type is occasionatlly found in affected lobules and groups of lobules, but is rarely found at the same stage in many lobules of the same group. More often the consolidation is less dense and the tissue is more resilient. This type of lesion is usually found accompanied by other types. (Fig. 8.)


FIG. 8.-Postinfluenzal pneumonia. Generalized confluent lobular pneumonia. Streptococcus hemolyticus was cultivated from the lung at necropsy. The pneumonia was accompanied by an acute hemorrhagic tracheitis, bronchitis and an acute mediastinitis. The histological picture was one of marked leucocytic infiltration of toe walls of the bronchioles, atria, sacculi, and alveoli, and an alveolar exudate composed of serum, little fibrin, and moderately numerous leucocytes, of which lymphocytes constituted about one-half. Lymphatics were distended with a similar exudate in which streptococci were abundant. There were relatively few organism, in the alveoli. Considerable areas of tissue were filled with extravasated blood. Accession number 3097, Army Medical Museum. Negative number 30666



A reaction, the greatest intensity of which is in the supporting tissue of the lung, has been recognized for some time as a part of the process in some cases of bronchopneumonia. Such a reaction is seen in the lungs of pneumonic plague, where it constitutes a prominent part of the picture. It is never the only type of reaction but lungs have been seen in pneumonic plague, and were described in the influenza of 1918, in which reactions other than inter stitial inflammation and its accompanying alveolar edema were not evident on macroscopical examination, though microscopical sections showed at least the early stages of consolidation.

The more acute stage was seen in cases dying in the first two to four days of illness. Grossly the lung resembled that of the acute fulminant type of reaction seen in the influenza pandemic but were more hemorrhagic, that is areas of frank hemorrhage were usual. (Plate I.)

Macroscopically such a lung is heavy and boggy. Though it does not collapse, crepitation is not easily elicited. On section the lung drips blood, and bloody serum which appears dull and brownish. The bronchi stand out, filled with hemorrhagic euxdate, the walls are thickened and often surrounded by hemorrhagic zones, while areas of frank hemorrhage already clotted are frequent. Histologically the peribronchial and perivascular tissues, aside from edema and hemorrhage, are infiltrated with lymphocytes, large mononuclear and polymorphonuclear cells, and the same infiltration is present in the alveolar walls. The alveoli are filled with a serous exudate which rarely contains fibrin. Streptococci are abundant in the bronchi and peribronchial and perivascular lymphatics but are seen in minimal numbers in the alveoli. They are present in the pleural exudate which often forms even in the rapidly fatal cases, presumably through infection spreading to the pleura from the hilus region.

Cases of interstitial pneumonia which live a longer period are accompanied by more involvement of the interstitial tissues and are relativley easy to diagnose macroscopically. The lungs of such cases are less distended and show areas of partial collapse. Areas of lobular pneumonia and of pneumonic consolidation extending from the peribronchial tissues, are always present. The bronchi show markedly thickened walls. The same type of thickening is present also along the vessels which frequently, except for the character of their lining, are indistinguishable from the bronchi. Fibrous tissue, separating lobules and groups of lobules which are involved in the process, shows marked thickening, is pale and opaque, the thickening being frequently of nodular character, due to irregular distention of the lymphatics with purulent exudate. (Plate II). All stages are seen from the acute hemorrhagic process, difficult to differentiate macroscopically, to advanced abscess formation in the interstitial tissues, accompanied by lobular and spreading bronchopneumonic consolidations, and empyema. In some instances the pneumonic consolidations are in excess, in others the interstitial reaction.


This is apparently the typical reaction of B. influenzae and is basically an involvement of the bronchi at their terminations in the respiratory bronchioles. At these terminal sites the infection extends to the alveoli grouped at the termi-



Streptococcus hemolyticus infection following influenza. Beginning consolidation. Accession 2696, Army Medical Museum. Colored photograph.



Secondary to confluent lobular pneumonia following influenza.
Accession 16646, Army Medical Museum. Colored photograph.


nation of the bronchial radicles producing minute miliary nodules, primarily hemorrhagic, later becoming necrotic and resembling tubercles either single or in groups, at times practically uniformly consolidating the lobule, the lesion appearing more advanced centrally. (See frontispiece.)

In the acute fulminant type, which is described under Camp Devens (p. 42), practically every bronchial termination is involved and the gross picture is that of hemorrhagic inflammation. Much more frequently fewer areas are involved in the early stage and other organisms determine the majority of the terminal pictures. In the peripheral portion of such lungs, however, the lesions at the ductus alveolaris are still evident usually presenting alveoli filled with leucocytes, many of which are the large mononuclear types, while proliferative changes are practically always present and may appear advanced even in cases which died within two weeks after the clinical onset. Metaplastic types of change are seen in the bronchial epithelium even in the fulminant types of a few days' duration.

Where the patients live a longer time the bronchial inflammatory reaction becomes frankly purulent. The process originally does not involve the entire lung but is progressive, involving one portion and then a succeeding one, so that in parts of the lung the hemorrhagic type of reaction is present, in others a purulent reaction in the bronchi with secondary thickening of the wall, a loss of epithelium and an extension by continuity for greater or lesser distances in the surrounding alveoli. In such cases, the smaller bronchi and surrounding infiltration microscopically resemble nodular tubercles. The pus in such cases usually contains very numerous influenza bacilli not infrequently mixed with Gram-positive cocci, of which the most frequent is the streptococcus. With a mixture of streptococcus in which the organism is found in the peribronchial tissue, it is difficult to decide which organism preceded, as either alone may result in an interstitial reaction, while bronchi are also blocked and organizing processes occur in both varieties of infection, perhaps somewhat more frequently and rapidly in cases due to influenza bacillus. However, there is a reasonable doubt as to whether the influenza bacillus itself produces a definite interstitial reaction unless the organizing processes in areas of atelectasis are so considered.

The definitions given above appear to be necessary in order to avoid confusion. These are based on the pathological anatomy rather than on the clinical aspects of the case, because it appears that there is general agreement neither between clinicians and pathologists nor among pathologists. There was wide variation in the physical signs which clinicians considered indicative of the type of pneumonia present. An abrupt onset, with pleural pain, particularly if dullness involved most of a lobar area, was generally termed lobar pneumonia. If a crisis occurred practically all would so diagnose it. Lobar is an anatomical term, however, and depends on the pathologic anatomy of the disease. A slow onset and a termination by lysis made little difference to the diagnostician, while to many pathologists the fact that a lobe was involved to a large extent, whatever the type or variation in type of the pathological process, was sufficient to cause him to designate it as lobar pneumonia. Many of the pathologists, however, stated that any type designation was not adequate, since practically all the cases showed various types, and several preferred the term pneumonitis or


inflammation of the lung to pneumonia, with the various qualifying terms commonly used.

Every type of acute respiratory lesion noted during the World War occurred during the influenza pandemic and in sufficient numbers to enable one roughly to classify them. The rapidly fatal type which was such a prominent and spectacular feature will be described first and illustrated, as it is considered that the lesions seen were but modified in extent, intensity, or both, in practically all cases to which the clinical diagnosis influenza was applied, while other lesions, except primary lobar pneumonia, were later complicating infections. Whether the fulminant cases were the result of the action of the bacillus of Pfeiffer or of this organism as a secondary invader, or of symbiotic action is still to be determined. The evidence even now, nine years after the pandemic, does not allow a definite answer to the important question of etiology.


With regard to the most numerous group-those who recovered in a few days-one may surmise a reaction less violent or less extensive than those seen in the fatal cases, but of similar type. Whether influenza is always primarily a respiratory disease or whether the infective agent may enter by other portals is not known. In the pandemic of 1918 the respiratory tract appeared always to be involved, at least in its upper portion.

Discussion of the lesions in these cases, in so far as these can be described from the Army material, follows:

Nose.- Grossly the redness, swelling, and mucous flow varied. In the earliest stages a sense of dryness and discomfort was followed by more or less profuse discharge of mucus. This might terminate the symptoms or more frequently a purulent inflammation followed. Accessory sinuses.- The ethmoids usually were involved if there were any marked lesion in the nose. The frontal sinus, maxillary antrum, and sphenoid frequently presented symptoms and at necropsy purulent inflammations of the mucous membrane lining them (one or several) were found, not infrequently accompanied by hemorrhage into the mucous membrane and exudate. The bacteriology is uncertain as so few were carefully examined. B. infIuenzae was found both early and late and the same is true for both pneumococci and streptococci. Extension of the process along the Eustachian canal and to the middle ear and involvement of the mastoid antrum in a purulent inflammation was observed with varying frequency. In the mastoid antrum bone necrosis often was observed while extension to the bones from the other sinuses was not frequent. It was more often seen in cases from which streptococcus was isolated. By bone involvement, extension occurred to the membranes of the brain both primarily and subsequent to operations on these sinuses. (Fig. 9.)

.- The mucous membrane here, so far as observed, was always reddened and the lymphatic structures swollen but not often to a great degree. No tissues were described.

Tonsils.- As a whole, little early change was seen except for hyperemia, especially in the crypts, which frequently showed red injected openings. The more pronounced changes seen in the cases of clinical influenza appear to have


been due to invasion by streptococcus and were not a part of the influenzal picture. The pillars of the fauces were reddened as was the uvula. Careful observation of the mucosa of the buccal cavity sometimes showed numerous reddened spots simulating an eruption; these were the inflamed openings of the buccal mucous glands.

Larynx.- The mucosa was swollen, reddened, and frequently showed smallerorlargerhemorrhagicinfiltrations of the superficial layers. Thecolumnar

FIG. 9.- Subdural abscess following suppurative frontal Sinusitis after influenza, Streptococcus hemolyticus. Accession number 12726, Army Medical Museum. Negative number 30804

cells were missing in some areas and the submucosa was edematous while the basement membrane was swollen and hyaline in appearance. On the surface small Gram-negative bacteria were seen and usually Gram-positive cocci having the morphology of pneumococci, streptococci, or both. In cases dying in the first few days the cellular exudate was often not abundant and contained a


large proportion of mononuclear cells. Occasionally, even early, the exudate along the surface was distinctly purulent and frequently contained numerous minute Gram-negative bacilli with a greater or lesser mixture of Gram-positive cocci.

Ulceration had not had the opportunity to form at this early stage and when present later it was probably due to organisms other than those of the primary infection. Atria of ingress, for whatever bacteria were present, were formed as a result of the denudation of the columnar cell layer.

.-The picture here was essentially the same as in the larynx. The mucous glands of the submucosa were more or less swollen and the ducts were filled with mucinous material containing mononuclear cells and occasionally frank pus. The gland cells appeared active and, as a rule, did not show marked degeneration or many bacteria in their vicinity. Rarely a few Gram-negative, small bacilli were found in the acini nearer the mucosa. Hemorrhages into the superficial mucosa were frequent and there were usually small areas filled with red blood corpuscles about smaller vessels in the submucous zone. The tissue between the ends of the cartilaginous tracheal rings was more or less infiltrated with mononuclear cells. Nodular accumulations of lymphocytes were reported occasionally as occurring in the submucosa and some of these showed the typical area of lymphocytic accumulation normally present at the bifurcation of bronchi within the lung, including a certain amount of reticular structure and reticulum cells. Ulceration was seen in patients who survived for some time but were not reported in the acute fulminant cases and were probably due in large part to secondary invasion. (Fig. 10.)

The tracheal lymph nodes were moderately swollen, moist, and red. The peripheral sinuses were dilated with fluid containing large mononuclear cells and lymphocytes while there was a general edema and an increased number of large mononjuclear cells in the tissue. In a considerable number of cases the enlargement of these nodes was extreme. Such nodes dripped blood and serum on section ahd hemorrhages were visible. The blood vessels of small caliber showed swelling and proliferation of the intimal endothelium and capillaries were often distended by hyaline "thrombi." The large mononuclear cells had often phagtcytosed leucocytes and occasionally contained red blood corpuscles, but it was unusual to find bacteria in their cytoplasm and when found the organisms were usually in phagocytosed polymorphonuclear leucocytes.

Proliferation of the reticulum cells, particularly those of the follicles, was usual and varied greatly in amount. It was most marked in cases surviving longest. (Fig. 11.)

.- On opening the chest the appearance of the thorax was rather characteristic. The lungs filled the pleural cavities and presented a large surface overlapping the heart, and decreasing the exposed area of the pericardium.

The loose tissues of the mediastinum were edematous and occasionally contained air. The pleural cavities contained a slight excess of fluid which was either clear or blood tinged.

The parietal pleura was more or less reddened and small hemorrhages into this membrane were observed. Definite pleural exudate in these fulminant cases was usually lacking. The lungs were a deep red or purplish color with


FIG. 10.- Hemorrhagic necrotic inflammation of the lining of the trachea in a fulminant case of acute, diffuse inflammation in both lungs. Both influenza bacillus and streptococcus were present in cultures. Accession number 3577 Army Medical Museum. Negative number 45910. Hematoxylin and eosin stain; X 21


FIG. 11.- Cellular reaction in a bronchial lympb node in pneumonia following influenza, showing the character of the cell exudate along the reticular tissue of the node. Very few polymorphonuclear leucocytes are seen, lymphocytes are moderately numerous, while large mononuclear cells preponderate; they were evidently actively proliferating as indicated by mitotic figures. Accession number 1049, Army Medical Museum. Negative number 45170.   Hematoxylin and eosin stain; X 580


FIG. 12.- Pneumonia following influenza. Diffuse, hemorrhagic inflammation of the lungs. Minute Gram-negative bacteria and streptococci in sections. The onset of pneumonia was four days prior to death. Pneumococcus, Group IV, found in the sputum, hemolytic streptococcus in the heart and lungs at necropsy. Accession number 2707, Army Medical Museum. Negative number 30602


occasional flecks or small extravasations of blood beneath the serosa. The lymph nodes at the hilus were large, moist, dark in color, and dripped blood on section; occasionally minute punctate hemorrhages could be seen. The lungs showed no firm consolidation but were boggy or rubbery in consistency, crepitation though present being elicited only in slight degree. The lungs were heavy and showed, even in necropsies made within a few hours after death, a greater density in the lower and posterior portions. (Fig. 12.)

Blisterlike structures often were seen on the pleural surface, more abundant anteriorly, but small ones were scattered over the entire surface in some cases. Some of these contained fluid but the majority were filled with air.

On sectioning such lungs, a great deal of fluid welled up on the cut surface. This fluid appeared to be blood and serum mixed, the blood content varying considerably so that from some parts of the lung, particularly the anterior portions, serum predominated, while from other parts the fluid appeared to be thin, dark blood. Large quantities of fluid escaped and more could be squeezed out as from a sponge, leaving a brownish-red tissue without definite consolidation. The fluid was frothy from air admixture and this was especially true of that which bubbled up in the severed bronchi, which usually contained less blood than that from the parenchyma. The tissue, as seen after wiping off the fluid with the knife, appeared firmer than normal, and with a lens the walls of the alveoli appeared thickened. In some cases hemorrhagic consolidations, appearing like small infarcts, could be made out surrounded by the hemorrhagic edema. At the time of necropsy these areas showed little evidence of inflammatory consolidation, but after the lungs had been fixed the center of these areas showed as paler nodules of peribronchiolar consolidation surrounded by zones of hemorrhage. (Fig. 13.) Emphysema was indicated by air vesicles protruding from the surface. Although there was some variation in the intensity of the reaction, the mucosa of all the bronchi was reddened and somewhat swollen, with a velvety sheen flecked with small hemorrhages.

Primary bronchi
.- The process being described appeared to extend along the respiratory passages with great rapidity. In a few hours an inflammation, apparently.starting in the nose or pharynx, would involve successively the larynx, trachea, and entire bronchial tree. Thus the bronchi showed the same type of changes seen in the trachea and because of the uniform involvement present throughouit the respiratory tract, little unchanged bronchial mucosa was seen in these fulminant cases at necropsy.

The mucus membrane, particularly of the larger bronchi, was usuallv intensely red, with a velvety sheen with occasional splashes or flecks of actual hemiorrhage. These tubes were filled with a frothy somewhat viscid mucus, tinged and streaked with blood, little exudate of a purulent nature being evident, though pus quickly appeared in the mitucuis if the patient survived for a sufficient time. Microscopically the columnar cells were swollen, many individual ones being cast off while here and there small areas of cells were raised from the basement membrane in small vesicles by fluid exudate. (Figs. 14 and 15.) It was unusual in cases dying in the first two to four days to find either grossly or microscopically complete denudation of very large areas. Suitably fixed speeimens show marked vascular engorgement particularly


of the capillaries just beneath the mucosa, while the lymphatic network in contact with the basement membrane was markedly dilated. The mucus glands appeared active and the nuclei as a whole stained well. Cellular infiltration was not marked. The majority of the infiltrating cells were of the lymphocytic type with rather numerous large mononuclear cells with pale cytoplasm, most abundant near the basement membrane.

FIG. 13.- Postinfluenzal pneumonia. Both lobes contain numerous disseminated nodules of consolidation, most of which are surrounded by zones of hemorrage. At necropsy the pale centers did not appear, being infiltrated with blood which tsasked the dense consolidation. The lesion in the left lung was a confluent lobular pneumonia superimposed on smaller foci of consolidation about the terminal bronchioles. Pneumococcus, Group IV, was isolated, but histological sections show minute Gram-negative bacilli. Acute tracheitis and bronchitis with flecks of hemorrhage into the mucosa. Accession number 2694, Army Medical Museum. Negative number 45879

In the well-fixed material minute Grain-negative bacteria were found, free along the mucosal surface, in the blisterlike areas and engulfed, sometimes in large numbers, by the large mononuelears. (Figs. 16, 17, 18, 19, 20, 91, and 22.) These organisms were not found deep in the tissues but were seen occasionally in phagocytes at 8 to 20 cells depth (50 to 200 u) beneath the surface. The higher up the bronchial tree sections were made, the more organ-


FIG. 14.- Early changes in bronchiolar epithelium. Cells show hypersecretion and beginning desquamation. Edema and cellular infiltration along the basement membrane and beneath it. Beginning purulent bronchitis. Accession number 1055, Army Medical Museum. Negative number 45195. Hematoxylin and eosin stain; X 600


FIG. 15.- Degeneration of bronchial mucosa in an acute fulminant case of pneumonia following influenza; necrosis of the cells of submucosal glands; hemorrhagic inflammation. Accession number 16595, Army Medical Museum. Negative number 45907. Hematoxylin and eosin stain; X 235


isms of the Gram-positive coccus type were found. These were rarely seen in the smaller tubes until later in the disease. Pus formed relatively early in the smaller bronchi and by the fifth or sixth day was usually abundant there. In tnis pus the Gram-negative bacilli were often present in enormous numbers, and as the disease continued the Gram-positive cocci increased in number as determined by stains of the tissue. (Figs. 23, 24, and 25.) These Gram- negative bacilli were not found as above described by all pathologists. They were present, however, in well-fixed tissue from all cases dying relatively early and subjected to necropsy soon after death which were examined in the studies on which this description is based.

Smaller bronchi
.- The changes continued of the same intense character as seen in the larger bronchi though with less regularity. Some bronchioles showed what appeared to be earlier, or at least less intense reactions, with less cellular response, less edema and desquamation and few bacilli. These structures of the smaller type were filled with exudate containing air or open spaces as seen in the sections. (Figs. 26 and 27.) There was less casting off of the mucosa and in many in which the inflammatory reaction was less intense there was a piling up of the cells from the basal layer with a loss of the columnar type of cell and an appearance of begininng metaplastic-like change toward the stratified squamous type, even in cases the duration of which was less than a week. (Figs. 28 and 29.) Consequently, these smaller bronchi were completely plugged by viscous secretion, the plug continuing down to their terminal branches and containing fibrin strands. The tissue distal to such a block was atelectatic.

The most characteristic changes in the pulmonary inflammation being described occurred in the true lung unit of Miller.2 This unit consists of the respiratory bronchiole opening into the ductuli alveolares which terminate in more or less spherical spaces beyond a dilated extremity. These spaces or atria lead from air sacs into which open the alveoli. In addition there are some alveoli along the respiratory bronchiole and ductuli which open directly into them.

The inflammatory reaction extends along the surface of the terminal bronchiole as a serous inflammation, the secretion filling the ductus alveolaris. It also extends into the alveoli leading off directly from the bronchiolar wall and to a greater or lesser extent into the atria and alveoli. In the fulminant cases many or most of these lung units are involved with a reaction varying from a serous to a hemorrhagic type accompanied by marked congestion and little cellular infiltration, giving rise to the edematous hemorrhagic picture seen macroscopically in the lung. (Pls. III and IV.) The distribution was rarely uniform but the more uniformly the lung as a whole was affected, the less varied the picture and the more rapid the death. Microscopically the ductuli alveolares were filled with viscid albuminous fluid containing air bubbles, while the atria contained more or less air, the exudate lying along the walls of the air sacs and alveoli. (Figs. 30 and 31.) The block by the exudate at the opening of the atria into the ductus alveolaris thus caused an increased pressure in the alveoli, a factor of undoubted relationship to the rupture of alveolar walls which so commonly occurred. The air sacs and alveoli opening from one ductus were rather


FIG. 16.- Gram-negative bacilli along bronchiolar epithelium in a case of lobular pneumonia. Streptococcus cultured from blood and lung. Duration eight days. Accession number 1519, Army Medical Museum. Negative number 45179. MacCallum stain; X 1070


FIG. 17.- Wall of bronchiole. Epithelium being raised and separated by serous exudate beneath. Gram-negative minute bacilli and mononuclear cells in exudate Accession number 3108, Army Medical Museum. Negative number 45270. MacCallum stain; X 1300


FIG. 18.- Lobular pneumonia. Hemolytic streptococcus in heart's blood, pleura, and lung. Pneumococcus. Type TV, regained at necropsy culture of lung. Minute Gram-negative bacillary and coccoid forms beneath the elevated mucosa of a bronchus. Accession number 3108. Army Medical Museum. Negative number 45198. MacCallum stain; X 1390


FIG 19.- Lobular pneumonia. Hemolytic streptococcus in heart’s blood, pleura, and lung. Pneumococcus, Type IV, in lung regained at necropsy culture. Gram-negative bacilli in macrophage cell beneath bronchial mucosa, which is raised by fluid exudate. Accession number 3108, Army Medical Museum. Negative number 45241. MacCallum stain; X1390


FIG. 20.- Minute Gram-negative bacteria in exudate of blisterlike elevation of bronchiolar mucosa. Accession number 3108, Army Medical Museum. Negative number 45237 MacCallum stain; X 1390


FIG. 21.-Minute Gram-negative bacteria, tree and in phagocytes in the wall of aductus alveolaris. Streptococcus hemoliticus recovered in cultures. Accession number 3101, Army Medical Museum. Negative number 45271. MacCallum stain; X 1390


FIG. 22.- Margin of ductus alveolaris, showing a short chain of streptococci and numerous minute Gram-negative bacteria. Post-mortem bacteriology showed streptococcus in the lung puncture, streptococcus, and pneumococcus in the pleural fluid. Accession number 3089, Army Medical Museum. Negative number 45326. MacCallum stain; X 3100


FIG. 23.- Beginning of ductus alveolaris. Minute (Gram-negative bacteria and streptococci in the exudate along the epithelium. Epithelial cells partially desquamated; basement membrane thickened and hyalin in appearance. The Gram-negative bacteria preponderated in exudate. Cultures from the lung and pleura showed pneumococcus, Group IV, and hemolytic streptococcus. Accession number 3l08, Army Medical Museum. Negative number 45199. MacCallum stain; X 1390


FIG. 24.- Gram-negative bacteria and a few streptococci along the wall of a ductus alveolaris. the epithelium of which is largely desquamated. Streptococcus hemolyticus in cultures from the lung at necropsy. Gram-negative bacteria were not cultivated. Accession number 1520, Army Medical Museum. Negative number 45188. MacCallum stain; X 1390


FIG. 25.- Wall of bronchiole infiltrated with leucocytes, Gram-negative bacteria and relatively numerous streptococcus forms. Streptococcus hemolyticus recovered in culture. Accession number 1520. Army Medical Museum. Negative number 45189. MacCallum stain; X 1390


FIG. 26.- Bronchiole in the center of a peribronchiolar consolidation. showing a seropurulent exudate, some proliferation of the basal cell layer with but little desquamation of the columnar cells. Basement membrane is thickened and the surrounding tissue edematous and infiltrated with leucocytes. “Hyalin thrombi” are seen in small vessels in the surrounding tissue. Accession number 2696, Army Medical Museum. Negative number 45999. Hematoxylin and eosin stain; X 235


FIG. 27.- Longitudinal section of a bronchiole containing purulent exudate in which is one large space formerly the location of a bubble of air. No organism isolated; organisms of streptococcic type in purulent exudate in the bronchi, Gram-negative bacteria in the ductuli. Surrounding lung tissue shows beginning consolidation, exudate being most dense near the bronchiolus from which it appears to have spread. Accession number 1049, Army Medical Museum. Negative number 45181. Hematoxylin and eosin stain; X 33


FIG. 28.- Bronchiole, showing purulent inflammation, proliferation of the basal layer and leucocytic infiltration. This is one of the larger bronchioles preceding the terminal divisions into the bronchioli respiratorii. Streptococci were present in the exudate in this bronchiole and Gram-negative bacilli in the peribronchiolar lesions of smaller size beyond it. Accession number 2694, Army Medical Museum. Negative number 46025. Hematoxylin and eosin stain; X 155


FIG. 29.- Respiratory bronchiole, showing purulent exudate which is continuous with that of the adjacent alveolar consolidation. A small group of alveoli open into the bronchiole on the right, where the wall is interrupted and the columnar cells merge with the flattened epithelium. So-called "hyaline thrombus'' in vein near bronchiole. The lymph spaces around the vein are edematous. Accession number 2696, Army Medical Museum. Negative number 46010. Hematoxylin and eosin stain; X 265


evenly involved and were often sharply defined by reason of the noninvolvement of the alveoli surrounding, and supplied by, other terminal bronchioles, though in the fulminant cases few of these terminal lung units escaped, most of them being filled with exudate mixed with air.

The exudate appeared to be viscid as it adhered along the alveolar walls as a homogeneous, eosin-staining, membranelike structure, the center of many alveoli being occupied by air. (Figs. 32 and 33.) In others a thinner serous exudate completed the filling of the alveoli as though the thin exudate had been thrown out early and was followed by the denser more viscid material which continued to adhere to the walls als a pseudomembrane. Beneath it the alveolar epithelium for the most part was lost, occasional cells and groups of cells remaining. Less of the viscid exudate was seen in alveoli the epithelium of which still remained. Fibrin stains revealed that there was little fibrin in this exudate, though occasional fibers were seen and the border of the membrane appeared more dense, stained more deeply with eosin, and reacted but slightly to the fibrin stain. The blood vessels of the alveolar wall were engorged with blood and there was escape of blood along the ductus alveolaris and about and into the alveoli. Rarely there was frank hemorrhage of considerable amount as occurred more often in lesions associated with streptococci. The hemorrhagic edema and universal vascular engorgement were responsible for the gross picture at necropsy.

The alveolar walls, due to edema and vascular engorgement, were prominent. Cellular infiltration, largely of cells of the lymphocyte type and large mononuclear cells from the tissue, was present in alveolar walls, and about the walls of the terminal bronchioles, while varying numbers were in the exudate. (Fig. 34 and Pl. V.) A frankly purulent exudate was rare in the fulminant case. The epithelium of the alveolar wall was lifted off by the exudate forming between it and the underlying reticulumn and appeared in the alveoli as single cells or small groups of cells. (Fig. 35.) The epithelium apparently formed a barrier to the egress of the exudate which, as stated above, appeared to be more abundant in alveoli, the epithelium of which had been cast off.

There are certain areas in the walls of the atria and alveoli where no capillaries are seen and special stains reveal no reticulum or very few fibers. Stains which differentiate elastic and collagen fibers reveal that such walls are made up of elastic and collagenous fibers on which rest the epithelial plates, there being a minimum of loose alveolar tissue between. The anatomical location of this type of wall suggests that there is a point of physical stress which has been reinforced by elastic fibers. These areas even in the earliest cases showed a swelling of the loose tissue beneath the epithelium which appeared to have no definite structure, and where epithelium was desquamated it merged with the exudate in the alveoli. Reticulum fibers were minimal or absent. Elastic fibers in their collagenous matrix remained as the only formed elements, the rest appearing necrotic. (Fig. 36.) At such places the ruptures of acute emphysema occurred. It would appear that though fibrous and elastic tissue was here in excess, a small vascular supply, possibly only a lymphatic supply, favored the necrosis of the tissue which, being at points of stress in the structure, yielded to the increased tension from exudate blocking plus respiratory effort. Whatever maty be the explanation, the ruptures of the walls of the air-containing tissue occurred most


FIG. 30.- Longitudinal section of a ductulus alveolaris which is less involved than the alveoli and air sacs surrounding it. It contains a serous exudate filling the atria and air sacs while the alveoli emptying into the air sacs and atria contain a few ells. The denser character of the exudate along the walls and in the termination of the mididle branch of the ductulus is apparent. The respiratory bronchiole is plugged with a dense exudate containing leucocytes. Such a plug prevented the emptying of this lung unit, and the respiratory efforts producing increased pressure were responsible for its dilatation. Accession number 3113, Army Medical Museum. Negative number 45559. Hematoxylin and eosin stain; X 85



Beginning purulent consolidation. Minute Gram-negative bacteria were seen in smaller air passages, a few streptococci were present in purulent exudate. Accession 2696, Army Medical Museum. H. & E. stain. Autochrome.



With beginning leucocytic infiltration of interlobular septum and alveolar walls, beginning streptococcus infection, many Gram-negative bacteria were seen along the bronchioles a few streptococci were present in bronchi and interstitial tissue.

Accession 2696, Army Medical Museum. H. & E. stain. Autochrome.


FIG. 31.- Cross section of an atrium, showing a few columnar cells at one corner, air and hemorrhagic exudate in the circular pale areas, and a dense viscid hyalin exudate (hyalin membrane), filling the rest of the structure and extending into the alveoli which open into it. The wall of the atrium is infiltrated with leucocytes, a considerable number of which are polymorphonuclear leucocytes. The primary reaction appears to have been an edema or practically universal involvement of the lung with a superimposed infection by streptococcus which produced the hemorrhagic inflammatory reaction. Streptococci were isolated and Gram-negative bacteria are seen along the respiratory bronchioles and in the atria. Accession number 2696, Army Medical Museum. Negative number 45960. Hematoxylin and eosin stain; X 235


FIG. 32.- Early exudate in a diffuse inflammation of the lungs in which lobular consolidation was commencing. Section taken from an area at the edge of the beginning consolidation shows a viscid exudate being thrown out from the alveolar wall and a thinner fluid filling the rest of the spaces. The exudate contains a few red corpuscles and a few leucocytes, both mononuclears and polymorphonuclears. The epithelium is largely desquamnated and a fews strands of fibrin are evident. This presents one stage in the formation of the so-called hyalin membrane. Accession number 3113, Army Medical Museum. Negative number 45580. Hematoxylin and eosin stain; X 230


FIG. 33.- Early exudate in a confluent lobular pneumonia, with formation of the so-called hyalin membrane about the walls of the atria and air sacs. The inflammation is in part hemorrhagic. Gram-negative bacteria and Gram-positive cocci with the morphology of pneumococci were found in this lung, but in the area illustrated Gram-negative bacteria predominated along the walls of the atria, pneumococcus forms in the exudate and in the surrounding interstitial tissue. Accession number 1385, Army Medical Museum. Negative number 45278. Hematoxylin and eosin stain; X 100


frequently at these points. Such ruptures caused emphysemnatous areas so that the atria and alveoli of one lung unit opened into adjacent ones. (P1. VI.)

Ruptures occurring near the interlobular septa caused escape of air into this tissue. The air then extended along the septa to the pleura, giving rise to air "blisters" beneath the pleura, and by extending mesially reached the peribronchial and perivascular areolar tissues. From the latter structures access to the mediastinal tissues occurred, there being no endothelial barrier. From the mediastinal tissues to the subcutaneous tissues the air extended along vascular trunks giving rise to the cutaneous emphysema. Air may also have escaped into the interstitial tissues from rupture of the bronchioles in the acute dilatations or bronchiectases which occurred in these structures, and escaped along this interstitial tissue to the hilus in the same manner as along the septa. Several observers considered this the chief method of escape of the air which in quantity gave rise to emphysema of the mediastinal and subcutaneous tissues.

The general picture just described, was not a frequent one, but occurred occasionally throughout the period of the war. It is not distinguishable grossly from the acute fulminant cases, where apparently Streptococcus hemolyticus was the etiological agent and, in many cases, the only organism found. In the lungs of cases dying within the first 10 days to 2 weeks, and often in those of longer duration, the picture described above was present in some portion of the lung, and probably represented an extension or a reinfection occurring during the clinical course of the illness. The study of many cases who succumbed later in the course of the disease revealed the subsequent stages in this process. In such cases the amount of pulmonary tissue affected by this type of reaction varied but was usually found unaccompanied by other processes in a relatively small proportion of the lung substance. An entire lobe or more than one might thus escape; more frequently the inflammation spread along all the main bronchi but did not involve all of the terminal bronchioles, so that the foci of inflammation were separated one from the other by intervening spaces of relatively normal tissue. In other words, the bronchitis was universal but the extension to the parenchyma was in scattered foci. (Fig. 37 and P1. VII.)

By studying portions of many lungs, it is possible to portray the various stages of the process. In the lungs of cases not dying during the stage of edema, the exudate gradually became more purulent; groups and single lung units became filled with pus in which polymorphonuclear leucocytes appeared in increasing numbers. While hemorrhage and engorgement predominated the foci were, macroscopically, small nodular areas, firmer than the surrounding air-containing tissue. The number of small Gram-negative bacilli increased and the inflammation extended to a limited extent into the alveoli and air sacs of hlng units adjacent to the original process. (Figs. 38, 39, and 40.) The filling up of the air sacs and alveoli with exudate gradually caused a change in the gross picture. The nodular foci became paler and less hemorrhagic and finally a reddish or yellowish-gray color resembling submiliary tubercles from which, macroscopically, they were practically indistinguishable. (Figs. 41, 42, 43, 43, 44, 45, and 46.) Fibrin appeared in relatively small quantities and lymphocytes and large mononuclear cells formed a prominent part of the cellular exudate. Some of the large mononuclears were actively phagocytic for leucocytes


FIG. 34.- Small arteriole between alveolar ducts, showing a ring of mononuclear cells which have apparently proliferated from the cells of the adventitia or arrived by migration through the vessel wall. The apparent attachment of some of them to the tissue suggests the adventitia of the vessel as the probable point of origin of these cells. Accession number 1049, Army Medical Museum. Negative number 45672. Weigert iron hematoxylin and picro-eosin stain; X 515


FIG. 35.- Early hemorrhagic lesion. Epithelium raised by underlying exudate from alveolar walls. Alveoli contain shadows of red corpuscles, a few lymphocytes and large mononuclears. Accession number 16571, Army Medical Museum. Negative number 45282. Hematoxylin and eosin stain; X 205



Hemorrhagic peribronchiolar pneumonia. Accession 1049, Army Medical Museum. Weigert hematoxylin, picro-eosin stain. Autochrome.



One of which involves the neighboring alveolus, the opposite wall of which is also ruptured. Accession 1519, Army Medical Museum. Elastic stain. Autochrome.


FIG. 36.- Degenerated wall of atrium. Swollen elastic tissue, stained dark, in a necrotic amorphous wall. Accesion number 1519, Army Medical Museum. Negative number 45221. Elastic stain; X 555


and occasionally contained red blood corpuscles. Occassional ones contained several nuclei either of nuclear division without cell division or formed as the result of the fusion of several cells. (Fig. 47.) The epithelium of the trachea and bronchi proliferated and appeared in multiple layers (figs. 48, 49, and 50), the cells approaching the squamous type in morphology and in arrangement, while the epithelial cells of some of the air sacs, particularly those not filled with purulent exudate, proliferated, were distinct, and appeared more numerous than normal. (Fig. 51.)

In some instances the stratified, squamous type of epithelium extended from the bronchioles into the alveoli or else the alveolar epithelium underwent this metaplastic-like change and filled the alveoli with solid masses of sells resembling foci of carcinoma. (Figs. 52 and 53.) If no secondary infection occurred to complicate the picture, the proliferative changes continued. Large mononuclear cells which previously had been seen about the adventitia of the arteries of the terminal bronchioles increased in number, and wandered through the exudate preceding the formation of reticulum which extended out from the walls of the atria, air sacs, and alveoli and formed a network in the exudate. (Figs. 54, 55, 56, 57, 58.) Collagenous fibrils were then laid down along and between the strands of reticulum which gradually disappeared except about vessels. (Fig. 59 and P1.IX; fig. 60 and P1.X.) In the early stages reticulum was laid down in irregular lines often beaded and also formed basket-like networks around individual cells which were morphologically large mononuclears, while collagenous fibrils were laid down in more or less parallel lines between fusiform cells of the fibroblast type. (Figs. 61, 62, and 63.) In addition to the reticulum formed in the exudate this material increased in the walls of the alveoli not always surrounding additional capillaries, though it is probable that, to some extent, these were formed later, the whole appearing as a sort of granulation tissue increasing the thickness of the walls of the atria, air sacs, and alveoli. (Figs. 64, 65, and 66.)

The organizing process as seen in cases living several weeks, usually dying from other infections, appeared grossly as indurated areas not definitely nodular, extending into the tissue somewhat diffusely and rather cyanotic, like recently formed scar tissue. This organizing pneumonia or organization of pneumonic exudate which occurred with extreme rapidity was characteristic of this inflammatory process. While this organization is seen most beautifully in the air sacs and alveoli, it also occurrred in the bronchi, though there it took the form of granulation tissue extending out into the lumens of the bronchioles from the capillary network beneath the basement membrane, no remnants of the latter being visible, thus indicating its complete destruction. (Figs. 67, 68, 69, and 70.)

The process described above in the terminal lung units was accompanied by similar changes in the bronchial tree. The bronchitis which early was of a catarrhal or serous type became purulent, the mucosa was desquamated from irregular areas and the infection extended into the peribronchial tissue. In the earliest stages the limiting membrane of the submucosa or basement layer became thickened and appeared hyaline and necrotic. The capillaries which terminated their loops at this membrane appeared increased in number. Ex-



Leucocytic infiltration of interlobular septa surrounded by hemorrhage. Minute Gram-negative bacteria and streptococcus forms were seen in the bronchioles, streptococci in the lymphatics of the septa.

Accession 2694, Army Medical Museum. H. & E. stain. Autochrome.


FIG. 37.- Peribronchiolar consolidation surrounded by zones of hemorrhage in the lower lobe. Peribronchiolar nodules in the lobules of tbe upper lobe surrounded by lobular consolidations due to Streptococcus hemolyticus Minute Gram-negative bacteria and streptococci in the bronchioles. Only streptococci recovered in culture. Accession number 3092, Army Medical Museum. Negative number 30716


FIG. 38.- Gram-negative bacillary forms in the alveolar wall with partial desquamation of the alveolar epithelium. Accession number 3108, Army Medical Museum. Negative number 45257. MacCallum stain; X 1150


FIG. 39.- Minute Gram-negative bacteria in the alveoli. Pneumococcus. Group IV, in lung culture. Hemolytic streptococcus in hearts blood, pleura, and lung. Accession number 3108, Army Medical Museum. Negative number 45252. MacCallum stain; X 1200


FIG 40. - Minute Gram-negative bacteria in alveolar wall which is infiltrated with lymphocytes, large mononuclear cells and a few Polyrnorphonuclear leucocytes. Streptococcas hemoloyticus found in culture. Accession number 3101, Army Medical Museum. Negative number 45233. MacCallum stain; X 1560


FIG. 41.- Peribronchiolar lesions in pneumonia following influenza. It is in this type of lesion that small Gram-negative bacilli most frequently are found. Streptococci were found in the bronchioles not far from these areas and in the interstitial tissue around the larger vessel seen in one corner of the illustration. Accession number 2694, Army Medical Museum. Negative number 45995. Hematoxylin and eosin stain; X 19


FIG. 42.- Pneumonia following influenza. Peribronchiolar nodular consolidations, with some spreading out of the lesion from the bronchioles so that a few lobules are filled with exudate. Considerable interstitial lymphangitis near the base of the lower lobe, with slight thickening of the interlobular septa and peribronchial tissue throughout the lung. Gram-negative organisms were present in sections of the peribronchiolar lesions; Streptococcus hemolyticus was cultivated from the areas of lymphangitis. Accession number 3036, Army Medical Museum. Negative number 42866


FIG. 43.- A bronchiole, showing absence of columnar cells at either side, the fibrinopurulent exudate being attached where epithelium is absent; the inflammation extends into the surrounding tissue. The places of attachment of the exudate were probably points where small groups or single alveoli opened into the bronchiole. Purulent infiltration of the submucous tissues. Streptococci and Gram-negative bacteria in the pus. This bronchiole was the center of a peribronchiolar, nodular consolidation. Accession number 2694, Army Medical Museum. Negative number 46028. Hematoxylin and eosin stain; X 170


FIG. 44.- Section of a ductus alveolaris near its termination, which is surrounded by purulent exudate in the alveoli adjacent. Two of these alveoli open directly into the ductus. Accession number 2694, Army Medical Museum. Negative number 46038. Hematoxylin and eosin stain; X 170


FIG. 45.- A small nodular consolidated area of peribronchiolar pneumonia. Section is taken through the division of a ductus respiratorius into atria. Atrial lumen filled with purulent exudate, as are the surrounding alveoli. The walls of alveoli adjacent to the tissue about the atria are swollen from inflammatory edema and are infiltrated with leucocytes. Accession number 2694, Army Medical Museum. Negative number 46032. Hernatoxylin and eosin stain: X 170


FIG. 46.- Consolidation about the division of a ductus alveolaris into its atria at about the termination of the ductus. Surrounding alveoli and air sacs filled with a purulent exudate. A small amount of columnar epithelium is seen at one end of the opening of the ductus. Some hemorrhage into the alveoli in the outer portions. The walls of the blood vessels are markedly thickened. Accession number 2694, Army Medical Museum. Negative number 46026. Hematoxylin and eosin stain; X 98


FIG. 47.- Infiltration of alveolar wall by mononuclear leucocytes, some of which have escaped into the alveolus. Large multinucleated phagocyte, which contains numerous leucocytes and some flecks of pigment. Accession number 22073, Army Medical Museum. Negative number 45565. Hematoxylin and eosin stain; X 650


FIG. 48.- Section of the trachea, showing complete loss of the columnar cells, with formation of a thick epithelial surface of stratified squamous type which extends down into the lumen of the duct of the tracheal glands. The latter show hypersecretion but little degeneration. Accession number 2694, Army Medical Museum. Negative number 46020. Hematoxylin and eosin stain; X 97


FIG. 49.- Section of trachea showing practically complete loss of columnar cells, with marked proliferation of the basal layer forming a mucosa which appears like the stratified squamous type. Inflammatory edema and leucocytic infiltration of the submucosa and glands with some degeneration of the latter and some hemorrhage from small capillaries. Accession number 2694, Army Medical Museum. Negative number 46043. Hematoxylin and eosin stain; X 85


FIG. 50.- Bronchus showing purulent exudate and air in the lumen. Columnar cells are completely desquamated and the basal layers are proliferated, forming an epithelium of the stratified squamous type. This is infiltrated with leucocytes. Surrounding tissues show inflammatory edema and leucocytic infiltration with an increase in lymphocytes in the node which shows at one corner of the figure. This type of reaction occurred early in the influenzal process. Accession number 2694. Army Medical Museum. Negative number 46029. Hematoxylin and eosin stain; X 145



Alveoli lined by columnar cells. Accession 22073, Army Medical Museum. H. & E. stain. Autochrome.


FIG. 51.- Area of atalectasis in influenza pneumonia. Proliferation of alveolar epithelium, connective tissue thickening of the alveolar walls, leucocytic infiltration of the walls with numerous leucocytes in the alveoli, most of which are large mononuclear types. Accession number 3038, Army Medical Museum. Negative number 45178. Hematoxylin and eosin stain; X 285


FIG. 52.- Section of wall of small bronchus, showing a loss of columnar cells, proliferation of basal layer into stratified squamous type of epithelium which extends along a small branch of the bronchiole and into an alveolus. No definite rupture of tissue is present, but in bronchiectasis breaks in the bronchiolar walls are usually covered with this type of epithelium. Accession number 16571, Army Medical Museum. Negative number 45273. Hematoxylin and eosin stain; X 142


FIG. 53.- Alveoli and atria more or less completely filled with masses of epithelium of the type of the basal or regenerative layer of the bronchi. The masses have the appearance of the stratified squamous epithelium of mucous membranes. This represents part of the process of organization in a peribronchiolar consolidation following influenza. Proliferative changes of this type are seen in the bronchioles and have extended into the atria and alveoli, or the epithelium of these structures has piled up and differentiated into the stratified squamous type. Accession number 16571, Army Medical Museum. Negative number 45272. Hematoxylin and eosin stain; X 131


FIG. 54.- Exudate in an atrium in an early stage of the process of organization. Some of the wandering cells have elongated nuclei which can be clearly made out in the albuminous matrix. This beginning of organization is taking place in an area where a hemorrhagic inflammation is present. Accession number 1049, Army Medical Museum. Negative number 45672. Hematoxylin and eosin stain; X 505


FIG. 55.- Wall of alveolus showing types of cells in the infiltration in an influenzal pneumonia near an area where organization is beginning. Duration 14 days. Delicate reticulum forms a basketry about some of the cells in the alveolar wall. Accession number 3572, Army Medical Museum. Negative number 45306. Reticulum stain; X 845


FIG. 56.- Early stage of organization of exudate in an alveolus. The reticulum is extending into the cellular mass and has formed a basketlike network about one of the cells. There is an increase in the number of capillaries causing a thickening of the alveolar wall. This same basketlike reticulum can also he seen about some of the cells in the alveolar wall. Accession number 3036, Army Medical Museum. Negative number 45668. Reticulum stain; X 810


FIG. 57.- Early stage of organization of the exudate in a pulmonary atrium. Reticulum is extending into alveolar exudate which contains relatively little fibrin. The fibers extend throughout the exudate and no apparent connection with cells can be seen. There is some regeneration of epithelium along the atrial wall. In the exudate are round and polygonal cells the exact character of which can not be determined except that there are a few lymphocytes. Accession number 3572, Army Medical Museum. Negative number 45488. Reticulum stain; X 745


FIG. 58.- Early organization. Mass of cellular exudate in an atrium, infiltrated with reticulum fibers, showing radii of attachment to the atrial wall; no evidence of formation of collagen. Accession number 1390, Army Medical Museum. Negative number 45553. Reticulum stain; X 375



Accession 3036. Army Medical Museum. Reticulum stain. Autochrome.


FIG. 59.- Organization of exudate in influenzal pneumonia. The central mass of organization shows dense fibers of reticulum stained black; the lighter lavender fibers are collagen which appears to be laid down between the reticulum fibers, which gradually disappear. In the cellular exudate in two of the alveoli reticulum fibers are extending tand in the advanced parts of the extending organization no collagenous fibers can be seen. Accesion number 3036, Army Medical Museum. Negative number 45671. Reticulum stain; X 250


FIG. 60.- Organization in a pulmonary atrium following influenzal pneumonia. Organized tissue has practically replaced the exudate. The reticulum fibers are stained black, collagenous fibers lavender. Accession number 1390, Army Medical Museum. Negative number 45552. Reticulum stain; X 250



Accesion 1390, Army Museum. Reticulum stain. Autochrome.


FIG. 61.- Organization complete in an alveolus, with fibers extending to a mass of cells in the atrium into which the alveolus opens. The cells are mononuclears, some of which have phagocyted pigment. This type of organizing tissue shows few reticulum fibers and much collagen. Accession number 22073A. Army Medical Museum. Negative number 45439. Hematoxylin and eosin stain; X 340


FIG. 62.- Higher power photograph of a portion ot the field shown in Fig. 61, collagenous fibers extendiog to the group of mononuclear cells. Accession number 22073B, Army Medical Museum. Negative number 45437. Hematoxylin and eosin stain; X 920


FIG. 63.- Practically complete organization of a group of alveoli, with the atrium in the center showing some vacant areas. The illustration shows the incomplete character of the organizing process. Inflammatory reaction is still present, as is indicated by the leucocytic infiltration. Accession number 1390, Army Medical Museum. Negative number 45307. Hematoxylin and eosin stain; X 250


FIG. 64.- Newly formed capillaries about a small blood vessel in tbe wall of a respiratory atrium, with increase in the number of blood vessels in the alveolar and atrial walls in the vicinity. Accession number 3036, Army Medical Museum . Negative number 45674. Reticulum stain; X 335


FIG. 65.- Increase in the number of capillaries in alveolar wzalls in the late stage of an interstitial inflammation of these structures, Accession number 3042, Army Medical Museum. Negative number 45564. Reticulum stain; X 255


FIG. 66.- Alveolar walls, showing new formation of connective tissue infiltrated with leucocytes. Duration 50 days. Accession number 3638, Army Medical Museum. Negative number 45183. Hematoxylin and eosin stain; X 210


FIG. 67.- Organization extending from an ulcerated area in a bronchiole into the seropurulent exudate in the lumen Very little fibrin was present, as shown by fibrin stain. Accession number 3042, Army Medical Museum. Negative number 45575. Reticulum stain; X 230


FIG. 68.- Granulation tissue in the wall of a bronchiole, projecting from an area previously ulcerated. It has been covered with a membrane made up of cells of the type of the basal layer of the mucosa. Columnar cells are not present. Accession number 3042, Army Medical Museum. Negative number 45584. Reticulum stain; X230


FIG. 69.- New fibrous tissue encroaehing on the lumen of a bronchiole in the process of organization. The surface is still free of epithelium but a small lumen remains. Blocks of bronchiolesof thischaracter were responsible for atelectatic areas beyond the point of block, and represent the results of the organizing process arising from the ulceration of the bronchiolar mucosa or extending in from alveoli which opened into respiratory bronchioles. Aceession number 3061, Army Medical Museum. Negative number 45193. Hemaloxylin and eosin stain; X 300


FIG. 70.- Late stage of interstitial reaction about bronchioles, showing the healing process well advanced, with overproduction of connective tissue, which, outside the immediate vicinity of the bronchi, is infiltrated with mononuclear leucocytes. The vascular submucosa is replaced with relatively avascular fibrous tissue. Accession number 3038B Army Medical Museum. Negative number 45202. Hematoxylin and eosin stain; X 210


tension to the air cells adjacent to the bronchioles, plus the thickening of the bronchiolar wall, caused these lesions to stand out prominently on the cut section of the lung as hemorrhagic to yellowish-gray opaque foci with a central opening, the lumen of the bronchiole, the color depending on the stage of the process at death. Proliferation of fibroblasts occurred at an early period of the process in the walls of bronchi and bronchioles.

The edematous bronchial wall thus was infiltrated with granulation tissue which gradually became dense connective tissue in an amount far in excess of that normally found. The infiltrating cells in such forming tissue were predominantly lymphocytes, which were sometimes very numerous.

The smaller bronchioles showed considerable distention in all stages of the process and frequently fusiform, and irregular dilatations or bronchiectases occurred and were evidenton the gross examination of the lung. Microscopically the dilated bronchioles usually showed complete desquamation of the columnar cells; the bundles of unstriped muscles were separated by more than the normal amount of intervening tissue which was edematous, the lymph spaces being widely dilated and the collagen fibers proportionately diminished. The hyaline layer of the basement membrane was interrupted and sometimes not visible in these dilated areas. The inflammatory reaction appeared more purulent, and the bronchial wall became infiltrated with leucocytes earlier than was the case of bronchioles not so affected. The alveoli adjacent to these bronchiectases were found in a more advanced stage of inflammatory reaction than those surrounding bronchioles whose walls were still relatively intact. Increased pressure was undoubtedly present in the terminal bronchioles due to the blocking of larger air passages by the viscid exudate of the disease plus respiratory effort. The focal character of these dilatations, however, is not readily explained. The presence of more advanced inflammatory reactions at these points may be either the cause or the result of the dilatation and injury of the tube.It is probable that the patchy distribution of the bronchiolar ulceration accounts for the focal character of the lesions. Proliferation of capillaries appeared to be unusually marked about bronchiectases and the fibrosis of the healing process was responsible for further distortion of the lumens of the bronchi and bronchioles. (Figs. 71, 72, 73, and 74.)


In few cases which died of pneumonia during the influenza epidemic and in relatively few deaths during the World War were the lesions in the lungs found to be of only one type. Pneumonic consolidation spreading out about bronchi was present in some part of the lungs in a large proportion of the fatal cases of acute respiratory disease. This lesion was associated with or produced by many different organisms; in fact, practically every organism isolated from the lungs was believed to have caused this reaction. Histologically, there were two forms of the condition both of which, however, were found associated in many lungs; these are described under the definitions.

Spreading bronchopneumonia, in so far as can be determined from the material reviewed, was a secondary type of pneumonia. In every instance where adequate histories were obtained there was abundant evidence of a preceding


disease most often definitely diagnosed either measles or influenza. The organism found in the tissues of such cases by examining microscopical slides was morphologically most frequently a pneumococcus. From the cultural standpoint pneumococcus Group IV was the most frequent organism, pneumococcus Type II atypical, streptococci of the nonhemolytic varieties, the hemolytic streptococcus and the type pneumococci, pneumococcus Type III being distinctly more frequent than Types I and II, were recorded as present in this type of lesion.

With the exception of the lesions in which streptococci were found, the sequence of events appeared to be first, a preliminary infection, most often bronchitis, as described under the fulminant cases in which the exudate in the bronchi became more purulent, pneumococci were found in addition to Gram-negative bacilli, the wall of the bronchus was infiltrated with leucocytes, the peribronchial area frequently contained fibrin in addition to serum, while strands of fibrin were often present in the purulent exudate in the lumen of the bronchi. (Plate IX; Figs. 75, 76, 77, 78, 79, 80, 81, 82 and 83.) Inflammation extended by continuity outwards into the adjacent alveoli apparently directly across the tissues and thence by the same process involved alveoli farther separated from the bronchus. In this way smaller bronchioles were involved as well as alveoli, so that the distal portions of bronchi supplying the air sacs in the vicinity of another bronchus might be involved from without, appearing from the microscopical section to have had the process extend to them along the peribronchial tissues, the lumen and the mucosa being affected last. (Fig. 84.) This caused: some confusion and resulted in the diagnosis of an interstitial type of reaction as the result of studying small microscopical sections when the actual process was an extension by continuity from purulent inflammation within another. bronchus.

Macroscopically, the lungs were more voluminous than normal, collapsing to a greater or less degree, depending upon the amount of tissue involved in the consolidation. Early in the process the periphery of the lung was sometimes free of involvement; more often it presented nodular areas which projected above the surface of the rest of the lung, the latter being somewhat collapsed, as the result of opening the chest cavity. The pleural surface of such areas usually was covered with a fibrinous exudate, which later in the process extended to the rest of the pleura and still later terminated in empyema, the organism present being that found in the tissue or, in lesions due to other organisms than the hemolytic streptococcus, that organism was usually found as well. Streptococcus also appeared alone in such cases, having reached the pleura and involved it possibly before the consolidation caused by other organisms had reached the periphery.

On section, the parts of lung affected by this process varied in their consistency. If caused by the pneumococcus they were usually firmer than normal, and sometimes distinctly dry, presenting reactions much like that of lobar pneumonia, though more frequently, as the result of there being less fibrin in the exudate, the consolidation was more elastic and less fragile. The more dense consolidation was near the bronchi and toward the hilus of the lung where several bronchi and their surrounding consolidations made one pneumonic mass.


FIG. 71.- Purulent bronchiolitis with bronchiectasis following influenza. Columnar cells are desquamated, and there, is considerable proliferation of the basal layer, forming a layer of cells of the stratified squamous type. Accession number 16571. Army Medical Museum. Negative number 45256. Hematoxylin and eosin stain; X 110