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 I




Measles in epidemic form made its appearance in our newly mobilized troops in the fall of 1917 and as the epidemic progressed the incidence of pneumonia as a complication increased just as it has in former wars. In most of the camps, in addition to measles, there was a high incidence of acute respiratory disease independent of measles, which assumed epidemic proportions and was accompanied by a large number of cases of pneumonia. At first the pneumonia corresponded with that seen in civil life, and this was particularly true in the northern camps. Mortality increased, however, in cases following the acute respiratory diseases and following measles, and it became evident that organisms of unusual virulence were present. Pneumococci of the fixed types showed their highest prevalence during the earlier weeks, becoming less frequent as the epidemic wave progressed, pneumococcus Group IV and streptococci replacing them in incidence in cultures. In certain camps where case fatality was high, bacteriological investigations revealed that hemolytic streptococci were present in a majority of the severe cases. Lobar pneumonia was seen in the early weeks of this outbreak of acute respiratory disease, but in typical form became much less frequent, being succeeded by lobular pneumonia, bronchopneumnonia or bronchiolar consolidations accompanying severe bronchitis. In pulmonary inflammations associated with hemolytic streptococci empyema was frequent, and many of the lungs showed a phlegmon of the insterstitial, peribronchial and perivascular tissues. This epidemic condition waned about the beginning of 1918.   

During the fall of 1917 the receipt of fresh troops was followed by a sharp increase in acute respiratory diseases, not only in the recent arrivals, but also among troops which already had passed through the acute respiratory epidemic. In the spring of 1918 in the majority of camps there was another epidemic wave of acute respiratory diseases, and this occurred whether or not any considerable number of new recruits had been received. At first the epidemic resembled that of the previous fall, and in the majority of camps continued to its termination without marked variation. In certain camps, however, there was a sharp rise in case fatality which, so far as can be ascertained from bacteriological examination, was due to hemolytic streptococci. Where there was found a relatively small proportion of these organisms in cultures of the sputum and lung, there was a relatively low case fatality rate. In a few camps where these organisms were prevalent, in the spring of 1918, the case fatality rate exceeded that recorded during the pandemic influenza of the following fall. In these instances the pandemic was not accompanied by a high incidence of hemolytic streptococci, but by nonhemolytic streptococci and pneumococci, a large proportion of the latter being Group IV.   

The morbidity of the three periods of high incidence of acute respiratory diseases of the World War varied,but the clinical description of the cases, the


pathology of the lesions and the bacteriology of the pneumonic processes were similar. Such variations as occurred in the bacteriology of pneumonia account for the differences in the case fatality rates.   

The acute respiratory diseases of the three periods appeared in three waves of influenza, the second having greater morbidity and mortality than the first, the third or pandemic one by far the greatest. The pandemic occurred at about the peak of the greatest mobilization in history of men into military units, a mobilization unavoidably accompanied by an interchange of the flora of the respiratory tract between the peoples of all great land divisions of the world, except South America.   

It is difficult to imagine a greater opportunity for enhancement of virulence by rapid passage from man to man of the organisms causing influenza, and of those producing secondary infections. Increase in virulence by this means must be considered as a factor of great importance in the production of the high morbidity and mortality.   

The pathological anatomy of the pulmonary lesions was characterized by its variety, but no new or hitherto undiscovered lesions were seen. The prevalence of but one type of lesion in the lungs of a single case was rare. Typical lobar pneumonia was infrequent, and when present was usually caused by pneumococcus Type I, occasionally Type II, and rarely by Group IV.

Bronchopneumonia, that is, pneumonic consolidation spreading out from the bronchial tree, was frequent, though rarely the only process present. This type of distribution resulted from extension of the inflammation from the bronchial lumen out into the parenchyma, largely by direct continuity. In some cases the inflammation spread to the walls of neighboring bronchi, enveloped them and, extending still farther, involved the surrounding parenchyma before it penetrated the bronchial wall. A similar gross picture was produced by extension outward from a lymphangitis or phlegmon of peribronchial lymphatics resulting from the drainage of pneumonic areas in pulmonary tissue not always supplied with air passages by that bronchus.   

The most frequent type of pneumonic consolidation was the lobular which occurred in isolated lobules in lungs, the major portion of which was affected by other types of distribution, and took the form of confluent or grouped lobular consolidations varying in extent from small areas to an entire lobe. In a lung where a large proportion of a lobe was involved, it was distinguished from lobar pneumonia by the lack of uniform hepatization, by variations in the character of the consolidation in different and often adjacent lobules, by its apparent extension by continuity at the periphery, and by its tendency to destroy pulmonary tissue. This resulted in necrosis, the formation of abscesses and gangrene. This type of distribution was the one usually present in lungs showing interstitial lymphangitis or phlegmon, caused by hemolytic streptococci.   

Acute bronchitis accompanied by small consolidations about respiratory bronchioles, and involving their terminal atria, air sacs, and alveoli, was found occasionally in the fall of 1917, more frequently in the spring of 1918, and almost constantly in some part of the lungs of the cases of the epidemic period of the fall of 1918. So far as can be judged by the examination of histologic sections, backed by a considerable amount of bacteriological evidence, this type of lesion


was associated with influenza bacilli, often in pure cultures, both in this country and abroad. In discrete lesions, the exudate was purulent, though the individual nodules frequently were surrounded by zones of hemorrhage. In some of the fulminant cases the nodules were definitely hemorrhagic and resembled small infarcts on gross examination, but in the microscopic sections these were seen to be peribronchiolar consolidations rather than hemorrhages about blocked vessels. This hemorrhagic type of lesion occurred in patients who died within the first 24 hours of the onset of pulmonary symptoms. None of the cases occurring in the troops in the United States died within 24 hours of the onset of respiratory symptoms, so far as can be learned from the necropsy records, though many died within 24 hours of the onset of symptoms indicating pulmonary involvement. The more rapid the fatal outcome after the onset of serious pulmonary symptoms, the more universal was the involvement of the lungs.   

In the early part of the pandemic wave of the fall of 1918, and occasionally during the epidemic of the spring of that year, cases occurred, the lungs of which showed a uniform lax consolidation. The lungs would sink in water, and were hemorrhagic, extremely moist, and dripped blood and frothy serum. Subsequent fixation in formaldehyde solution usually revealed areas of early consolidation and zones of clotted hemorrhagic infiltration. In such cases it appeared microscopically as though practically every alveolus in the lung was uniformly and simultaneously involved with a serous inflammation completely filling the lung with fluid which was accompanied by varying amounts of hemorrhage. In these cases death probably was from suffocation rather than the toxic effects of the etiologic agents. Such cases were responsible for the feeling that the old classification of pneumonia by distribution of the lesions was inappropriate, and that the name acute pneumonitis was more applicable. In such cases bacteriologists found influenza bacilli, and in the collections of the Army Medical Museum every case examined of this type in which the tissues were well fixed, showed small Gram-negative bacilli, though all of this type came from cases dying during the pandemic of the fall of 1918. Lesions of this character involving lobes or parts of lobes were seen frequently in cases of less fulminant character, and represented extensions of the process shortly preceding the death of the individual. All variations in the confluence of the process were seen, from the isolated foci, which resembled conglomerate tubercles, to uniform generalized pneumonitis.   

In some instances hemolytic streptococci invaded the lung and produced a fatal issue within the first few days. On gross examination these lungs showed more hemorrhage than when the influenza bacillus alone was present, but the most striking difference was seen in the microscopical examination which revealed as the characteristic picture a leucocytic infiltration of the alveolar walls. In some cases this infiltration was predominantly lymphocytic, but usually polymorphonuclear cells equaled or exceeded the lymphocytes.   

Mononuclear leucocytes were seen in lungs of the diffuse pneumonitis type from which influenza bacilli were cultivated, but the principal changes in the alveolar walls of such cases were edema, greater or lesser desquamation of the epithelium and, in many instances, the formation of the so-called hyaline


membrane. This membrane was formed by a secondary exudation of more viscid character along the alveolar wall subsequent to the casting off of the epithelium of the alveoli and atria. It was rarely seen along walls still covered with epithelium.   

The picture at necropsy, except for the variation in distribution, depended to a very large extent on the length of the time the patient lived following the onset of the pulmonary complications. The earliest deaths were due to diffuse inflammations. Later came the various types of distribution of the consolidation, streptococcic lymphangitis occurring sometimes apparently simultaneously with the consolidation, or, where this lymphangitis occurred in lobar consolidations, it usually manifested itself during the stage of gray hepatization. Interstitial lymphangitis and phlegmon were secondary to streptococcic inflammation of the alveoli and bronchi and formed a prominent part of the picture in most of the cases caused by streptococci. Unlike the pneumococcus, the hemolytic streptococcus spreads through the lymph spaces of the alveolar walls rather than along air passages, irritating the lymphatic channels which drain the regions invaded by it, so that they take part in the inflammation. The lymph channels become blocked by the clotting of the fibrinopurulent exudate and in them abscesses are formed. It is also possible that when lymphatic channels at the hilis are blocked by such a process the lymphatic circulation may extend outward into the lung by collaterals as yet uninvolved; but this method was not frequent. The lymphangitis usually was secondary to the inflammation of the air passages and alveoli.   

Necrosis and the extension of peribronchiolar lesions to complete involvement of the lobules required that the patient live a longer time. With increased length of life hemorrhagic features disappeared. The lesions were first purulent, and later there occurred necrosis, which formed so prominent a feature in cases dying after two or three weeks of illness. Abscesses formed slowly except in pneumonic lesions due to staphylococci in which the rapid lytic action of this organism gave rise to definite abscessed lesions about the bronchioles in cases dying within a week of the onset of pneumonic symptoms.   

Numerous complicating or associated lesions of the various types of pulmonary inflammation have been described. Certain of them appeared with such frequency or were of such nature as to excite considerable interest. Acute emphysema, giving rise to blebs beneath the pleural surface, air in the interstitial tissues of the lung, in the mediastinum, along vessels, and in the subcutaneous tissues, was a spectacular condition. Escaping as the result of degeneration and rupture of the alveolar and atrial walls, possibly also through ruptures of acute bronchiectases of the bronchioles, it made its way along interstitial tissues, perivascular and peribronchial structures, and eventually appeared in the subcutaneous regions. Occurring as it did when the medical profession was interested in wounds infected with the Welch bacillus, it caused considerable concern but was of little clinical importance. Much more important was the extension of the inflammatory process to the pleura, usually after considerable portions of the lung had been involved, but occasionally directly from the mediastinuim without much, if any, pulmonary involvement other than an


acute bronchitis. The organisms responsible in most instances were hemolytic streptococci.

Empyema is adequately described in another volume so its discussion in this is limited solely to tracing its relationship to the other pathology of the respiratory tract.   

The involvement of the peritoneum, of the meninges, and of the joints, other than from direct extension from other processes, was probably due to the virulence of the invading organism, and though infrequent, occasionally occurred in every camp. Organisms isolated from these lesions were occasionally streptococci, more frequently pneumococcus Type II, occasionally Group IV. In several camps pneumococcus Type II showed a virulence exceeding that of all other organisms.   

Atelectatic areas in the lung were caused by the blocking of the small bronchi by the primary inflammatory processes, also as a result of abscess formation which encroached upon the lumens of bronchioles.   

Thickening of the epithelium of the air passages by a proliferation of the basal cells after the columnar layer had been desquamated was seen even in the earliest cases, and was particularly noticeable in those from which influenzal bacilli were cultivated. The result of this proliferative change resembled stratified squamous epithelium and sometimes extended out from the bronchi into the alveoli, particularly those alveoli situated along the walls of the respiratory bronchioles. Some part of the respiratory epithelium in every case showed this change in varying degree, but it was absent in the trachea and bronchi of cases in which streptococcus infection apparently occurred early, where usually the surface epithelium and sometimes the hyaline basement layer was destroyed.


The diagnosis "influenza" was used throughout the history of this camp. There was a definite epidemic outbreak during the months of March and April of 1918, although the most marked incidence was in September and October, 1918. Measles was a serious factor at this camp and was responsible for a considerable number of deaths, showing a sharp increase at the time of the original mobilization, and just before and just after the influenza epidemic, although, due to a decrease in personnel, the high rates of November and December, 1918, do not represent very many cases.   

The exact character of the pneumonia preceding the epidemic of influenza is not indicated in the history of the camp or in reports on the subject. It is probable that a certain number of the cases of primary pneumonia, as in other camups, were secondary to a preceding acute respiratory infection. There is no indication from available records that empyema was a particularly serious complication in this camp, and the high percentage of cases diagnosed lobar pneumonia strongly suggests a preponderance of the fixed types of pneumococci both in the fall of 1917 and the spring of 1918.b

Source of information, except as otherwise indicated: Medical reports to the Surgeon General. 1917, 1918, and 1919.

b The following statements of fact are based, in the main, on: A Review of the Epidemic of Influenza at the Base Hospital, Camp Beauregard, La., with Special Reference to Symptoms and Sequelae, by Donald J. Frick, American Journal of the Medical Sciences, Philadelphia, 1919, clviii, 68-80.

12 CHART I.- The incidence and fatality of the acute respiratory diseases at Camp Beauregard

13   The symptoms were of exceeding variety as to severity and kind. Most of the cases had severe backache, headache, a racking cough, flushed face, injected eyes, and, as a rule, pulse between 80 and 100, with a temperature from 101° to 104° F. Most of them were dull and apathetic, went to sleep in the ambulance, slept on the litter while they were being carried into the hospital, and had to be aroused to answer questions. They roused up only to cough, take medicine or nourishment. All looked prostrated and sick. All slept 12 hours, some 16 to 24 hours. The fever and prostration in the mild cases lasted 48 hours; the patients then began to get up and sit on the edge of the beds. They were all weak and some fainted in trying to get out of bed. The more severe cases had fever for three or four days. Practically all cases that had fever longer than this proved to have bronchopneumonia. Many had vomiting; some became tender over the abdomen, imitating an intra-abdominal condition.  
There was a leucopenia in practically all cases and no rise of leucocytes with the advent of pneumonia.   

Besides the ordinary symptoms, a striking feature in the early stages of these cases was the bleeding from some portion of the body; 15 percent suffered from epistaxis. Everywhere one went one saw men with bleeding or packed noses. Fifty cases of subconjunctival hemorrhage were counted. Twelve had a true hemoptysis, bright red blood with no admixture of mucus. Six cases vomited blood. One case died from loss of blood from this cause. Among the nurses, menstruation began in about 80 percent 2 or 3 days after going to bed. Three cases had intestinal hemorrhage. Most of the bleeding came early, with the exception of three cases of hemoptysis, which came after bronchopneumonia had begun.   

Twelve to 15 percent of all cases had herpes either of the lips, tongue, nose, or mouth, the eruption being exceedingly extensive in some cases.   

Pneumonia developed secondarily to the influenza in 1,342 cases, and the complications given below must be regarded as complications of this disease, in most cases, rather than of influenza.  
Jaundice of a mild or severe degree developed in 112 cases of bronchopneumonia. In those recovering, the jaundice disappeared very rapidly, in some cases clearing up entirely in seven or eight days. The liver in all cases seen was enlarged, so it was believed to be a true cholangitis.   

Otitis media was reported in 41 cases. Fortunately the otologists were on duty day and night and did immediate paracentesis on all bulging drums. No mastoid operations were necessary.   

Infectious psychoses were seen in 18 cases, the patients exhibiting all grades of aberration from simple transient hallucinations to maniacal frenzy which needed mechanical restraint for 12 to 24 hours.   

Two cases developed an hysterical aphonia (which cleared up in three and eight days respectively).
Five cases of pneumococcus meningitis were diagnosed clinically and proved positive by lumbar puncture; six to eight cases were punctured that had some signs of meningitis; these all showed a clear fluid under pressure, no increase in cells and no bacteria.


Eleven cases of epidemic cerebrospinal meningitis were diagnosed and isolated during the epidemic; three cases developed after the patients were convalescing from pneumonia and had been in the hospital from three to four weeks.   

Empyema cases appeared at the end of the epidemic. Twenty-nine were reported, 3 cases showed hemolytic streptococcus, 20 cases pneumococci, 7 cases no organism found.   

The most interesting complication on account of its usual rarity was interstitial emphysema, of which there were 10 cases. The only reason that can be given for the great number of these cases is the violence of the cough and the degenerative changes brought about in the disease. All of these cases had influenza, followed by bronchopneumonia, the subcutaneous emphysema coming from 6 to 10 days after the beginning of the disease. The air was first felt in all cases in a collar about the neck and then gradually spread over the back, chest, and arms to the hands as far as the knuckles and down the thighs to the middle. A rupture of the bronchus at the root of the lung seemed to be the only explanation.  
Two cases after violent coughing developed a pneumothorax, the heart being pushed to the right as far as the midline on that side. One case proved to be a pyopneumothorax.   

No reason could be given for the frequency of hemorrhages, but it was felt that there must have been some definite change in the constituents of the blood, although no estimation was made of the coagulation and bleeding time, nor were the platelets counted.   

Death came rapidly and could be prognosticated practically from the moment the man entered the hospital.
Autopsy protocols of some of the most interesting cases are given below.

AUTOPSY PROTOCOL, No. 1 (Autopsy No. 6)   

E. V., Sergt., Co. E, 5th Infantry. Time of death: 2 a. m., September 28. 1918. Time of autopsy: 9 a. m., September 28, 1918. Admitted: September 26, 1918.  

Clinical diagnosis
.- Lobar pneumonia, right lower. Anatomical diagnosis.- (1) Acute catarrhal bronchitis; (2) diffuse bronchopneumonia of both lower lobes; (3) acute hemorrhagic pleuritis of right lung; (4) acute splenic tumor; (5) cloudy swelling of liver, kidneys, and myocardium. External examination. - The body is that of a well-developed, white male, about 71 inches long, weighing about 185 pounds. Strong rigor is present, and there is marked lividity of the posterior surface of the body and of the left side of the face. There are no external marks of injury or of identification. Internal examination.- A thin layer of subcutaneous fat is present. The peritoneal cavitv is free of adhesions and of fluid. The pericardial cavity contains a slight excess of fluid which is partly clotted. The left pleural cavity is normal. The right contains a slight excess of fluid, which is faintly clouded and slighly blood-tinged.   Lungs.- The pleural surface of the left is smooth and glistening. The lower lobe, although it contains air, feels heavier and less crepitant than normal. On section the lower lobe is congested, and the cut surface, from which frothy fluid escapes, is very finely granular, the appearance suggesting the consolidation of innumerable groups of small numbers of alveoli, these being separated by small groups of air-containing alveoli. The upper left lobe is crepitant and pale pink in color. The pleura of the lateral surface of the right lung is finely granular and hemorrhagic, but its surface contains fibrin. The lower lobe is con-


gested, and like the lower left lobe is heavier and less crepitant than normal. On section it is somewhat more deeply congested than the lower left lobe, but has the same very finely granular appearance. The upper and middle left lobes appear normal. The bronchi of both lungs are deeply congested, but contain no exudate.  

.- The right side is distended with blood. All the valves and the root of the aorta are normal. The heart muscle is pale and cloudy.   Spleen.- The spleen is enlarged to about three times the normal and feels rather soft. The outer surface is smooth and is dark bluish red in color. On section the tissue is dark red and much blood escapes from the surface. Liver.- Normal externally. On section the tissue is cloudy. Gall-bladder and ducts are normal.   Kidneys.- Each is slightly increased in size. On section the cortex is swollen and cloudy and the glomeruli stand out as minute, slightly congested points.   Adrenals, pancreas, gastrointestional tract, bladder, and prostate.- Negative.   Bacteriology.- Cultures from the heart's blood and from the fluid of the right pleura show no growth. Smears from the lower lobes of the lungs show Gram-negative small bacilli and Gram-positive diplococci. Cultures from the lung yield B. influenzae and a Grain-positive, bile insoluble, nonhemolytic streptococcus.


F. W. T., Pvt., Co. E., Dev. Bn. Admitted: September 29, 1918. Died: 9 a. m., October 4, 1918. Autopsy: 10 a. m., October 4, 1918.  

   Clinical diagnosis
.- Lobar pneumonia.    Anatomical diagnosis.- (1) Lobar pneumonia of the right lower lobe (red hepatization); (2) pseudolobar pneumonia of the left lower lobe; (3) bronchopneumonia of both upper lobes and of the right middle lobe; (4) acute catarrhal bronchitis of both lungs; (5) acute fibrinopurulent pleuritis of right lung; (6) acute splenic tumor; (7) cloudy swelling of liver, myocardium, and kidneys.    External examination.- The body is that of a slenderly built white male, about 70 inches tall, weighing about 146 pounds. The body is still warm and rigor is absent. The neck and face are cyanotic. On the inner surface of the left knee is a large old sear of a burn.    Internal examination.- Peritoneal, pericardial, and left pleural cavity are normal. A thin layer of yellow exudate is present over the lower outer half of the right upper lobe and between the lobes.
- The lower lobe of the left lung is occupied by irregularly shaped confluent patches of dark red, airless tissue, which is congested and granular on section. A smaller wedge-shaped area is present at the lower border of the right upper lobe. The right lower lobe is completely consolidated and on section the tissue is dark red and granular and contains much blood. A few small patches of consolidation are present in the right upper and middle lobes. The bronchi of both lungs are congested and contain no exudate.   Heart.- Negative except for cloudy swelling of the muscles.   Spleen.- The spleen is increased to twice the normal size. On section the tissue is dark red and soft.   Liver.- Cloudy.   Kidneys.- The cortex is swollen and cloudy.   Adrenals, pancreas, gastrointestinal tract, bladder, and prostate.- Negative.   Bacteriology.- Smears from the right upper and lower lobes and from the pleural exudate contain many Gram-positive diplococci. Cultures from these situations contain pneumnococci. Culture from the spleen is negative. The heart's blood culture is contaminated by a large Gram-negative bacillus, but contains Gram-positive diplococci, which prove to be Type IV pneumococci in the peritoneal exudate of a mouse injected with the culture.


R. F., Pvt., Co. H, 5th Infantry. Admitted: September 26, 1918. Died: 12.50 p. m., October 10, 1918. Autopsy: 3.30 p. m., October 10, 1918.  

Clinical diagnosis.- Influenza. Bronchopneumonia of the left lower lobe. Interstitial emphysema bilateral.


Anatomical diagnosis
.- Interstitial emphysema bilateral. (1) Pseudolobar pneumonia of both lower lobes; (2) bronchopneumonia of both upper lobes and of the right middle lobe; (3) interstitial emphysema of the mediastinum and of the subcutaneous tissue of the face, neck, thorax, and arms.    External examination.- The body is that of a sparely built white male about 66 inches tall, weighing about 130 pounds. Rigor and lividity are not present. The skin of the right side of the face and of the entire neck appears swollen, and on pressure it has a crepitant crackling feel. The same crepitant sensation is elicited upon pressure of the skin of the chest anteriorly as far down as the nipple line, literally on both sides of the chest to a slightly lower level and about halfway down on both arms. The skin of the crepitant regions is not discolored.   Internal examination.- The peritoneal cavity is normal. When the sternum is removed the tissues of the mediastinum are found filled with air, which occurs in the form of bubbles of varying size held in the loose tissues. The loose tissues of the root of the neck show a similar condition. The pericardial cavity is normal. Both pleural cavities are free of fluid and of adhesions. The lungs are not collapsed and no air is present in the pleural cavities. Lungs.- The two lungs are removed in toto, together with the trachea. Bubbles of air are present beneath the pleura of the left lung; these bubbles follow the lines of the interlobular septa. The pleura of each lower lobe posteriorly and laterally is granular and cloudy, but no macroscopic fibrin is present. All the loose tissue about the roots of the lungs contain bubbles of air. The lungs feel heavy, the lower lobes being enlarged and dark red in color. Areas of consolidation are present in both upper lobes, and about half of the right middle lobe is solid. Although the lower lobes are almost completely consolidated, small areas of air-containing tissues are present at the margins and in several places beneath the pleura. When the lungs are immersed in water and inflated by a tube placed in the trachea a steady stream of air bubbles escape from the tissue at the bifurcation of the trachea, the air apparently coming from the left bronchus anteriorly at the level of the bifurcation. On opening the trachea a small tear is found in the mucosa of the anterior wall of the left bronchus in the region from which the air was seen to escape. On section of the lungs both lower lobes are dark red and granular. Scattered about in the dark red tissue are numerous more opaque gray areas two or three lobules in size; these areas appear to have a peribronchial distribution. The right middle and both upper lobes contain areas of consolidation, of varying size, which are grayish red in color and rather soft and friable. The bronchi contain no exudate and their mucosa is not reddened.     Spleen.- The spleen is moderately enlarged, firm, and dark red. The remaining organs were not examined further.    Bacteriology.- Smears and cultures from the right lower lobe show nothing. Cultures from the spleen and heart's blood remain sterile. Smears from the left lower lobe show a few Gram-positive diplococci; the cultures show numerous Gram-negative small bacilli and Gram-positive cocci in pairs and chains. The latter type are Type III pneumococcus.


W. McC., Pvt. Co. A., M. P. Died: 5.15 a. m., October 20, 1918. Autopsy: 11 a. m., October 20, 1918.  

  Clinical diagnosis
.- Pneumothorax.   Anatomical diagnosis.- Pneumothorax (left); suppurative pneumonia of the left lung; bronchopneumonia in the right. Histological section of the tissue of the right lung, left lung, and spleen.   External examination.- The body is that of a well-developed male, weight about 165 pounds, 6 feet in height. Rigor mortis is present. The skin is clear and there are no deformities or marks; the appearance of his face indicates some emaciation.   Internal examination.- An incision is made from the suprasternal notch to the pubes. The subcutaneous tissue is in good shape.

Lungs.- The left lung has completely collapsed. It is nodular, grayish in color which is due to a fibrinous exudate, but there are no adhesions and a little fluid containing pus in the pleural cavity. On section pus flows freely from every nodule. The lung is full of abscesses


ranging from the size of a hen's egg to minute abscesses. The bronchi are filled with pus. The right lung has a consolidation in the lower posterior portion of the upper lobe involving the upper posterior third of the lower lobe. The rest of the lung feels as though it is in fairly good shape, with plenty of air in it. On section there is pus in the consolidated portions and also some little pus in the bronchi. The air cells in the seemingly good portions of the lung contain in various areas a little frothy fluid but there are no evidences of inflammation. There are no adhesions except the adjacent consolidated portion. There is little fluid in the right pleural cavity.    Heart.- The heart is negative.

The liver, pancreas, intestines, and peritoneum are all negative. Spleen.- The spleen has an infarct about the size of a 50-cent piece on the upper anterior third. On section it seems inflamed and is somewhat enlarged. Kidney.- The right kidney is large, red, and seems swollen, and the capsule peels off easily, but on section it is negative. The left kidney seems identical with the right. The appendix is negative.

The bacteriology and pathology of the epidemic were given special consideration.c   

On the first day of the epidemic it was decided to select two widely separated wards and to make nose and throat cultures on every patient. Half of the cultures were taken on whole blood-agar plates and half on laked blood-agar plates. One plate was used to a patient, the nasal swab being smeared over one half and the throat swab over the other half. On the same day specimens of sputum from patients with productive coughs were examined microscopically and washed portions injected intraperitoneally into mice. Next, six patients who had the highest temperatures were selected as subjects for blood cultures, 10 c.c. of blood being inoculated into 150 c.c. of plain meat-infusion broth.
The examination of the nose and throat cultures the following day showed that the vast majority of the plates contained countless numbers of small colonies transparent to transmitted light, and of such minute size that often the aid of a hand lens was required for identification. Smears from these colonies showed large numbers of small Gram-negative bacilli having in general the morphology and staining characteristics of the B. influenzae, but showing considerable pleomorphism. Colonies were fished and subcultures made and the same organism obtained in pure culture. The organism proved to be a very delicate one, and even when subcultured daily on blood medium soon died out. It was observed that while large numbers of colonies were obtained from both nose and throat, the greater numbers were found in the throat cultures. In one ward 35 patients were cultured and everyone showed positive cultures as described above. Three of the plates also showed hemolytic streptococcus colonies. Unfortunately the cultures from the other ward were not reported on exactly, as part of this record was lost. However, the findings were practically identical. The influenza colonies grew to much larger size when in the vicinity of staphylococcus colonies. In the second ward there were fount five hemolytic streptococcus carriers.   

The direct microscopic examination of the specimens of sputum showed both small Gram-negative bacilli and Gram-positive diplococci, the latter

cThe following statements of fact are based, in the main, on: Bacteriological Observations of the Epidemic of Influenza at Camp Beauregard, La., by J. E. McClelland. American Journal of the Medical Sciences, Philadelphia, 1919, clviii, 80-87.


predominating. From the peritoneal exudate and the heart's blood of the inoculated mice both B. influenzae and Group IV pneumococci were obtained.   

All the six blood cultures taken on the first day of the epidemic remained sterile, nor was it possible subsequently to identify B. inftuenzae in any of the blood cultures.  
It was then decided to make throat cultures of the entire hospital, as far as practicable, with the idea of determining how many patients would show typical influenza colonies, and also in order to locate the heinolytic streptococcus carriers, that they might be isolated from the other patients. From September 25 to October 3, throat cultures were made on 1,919 patients. Of these 1,749, or 91.1 percent, were positive for influenza, 80 were positive for hemolytic streptococcus, 39 showed both influenza and hemolytic streptococcus and 90 were negative. The total number of hemolytic streptococcus carriers then was109, or 5.7 percent of the entire number cultured.  
As soon as the cases of pneumonia developed the attempt was made to make microscopic and bacteriological examinations on as many of the sputa as possible. While many of the direct smears showed numerous Gram-negative bacilli, the predominating organisms were Gram-positive diplococci having the morphology of pneumococci. Type determinations for pneumococci were made on 723 cases, using the Avery method. The results of type determinations were Type I, 1.1 percent; Type II, 8.4 percent; Type III, 11.6 percent; Group IV, 78.9 per cent.   

From a consideration of the relative incidence of the various types in normal carriers it would seem reasonable to suppose that many of the patients developed a pneumonia that was really an autoinfection from the organisms already present in their own upper passages. Undoubtedly the added influenza infection played a very important role either in reducing the patients' general or local resistance to secondary infection, or in some way enhancing the invasiveness of those organisms which were shown to be the most important etiological factors in the complicating pneumonias. At any rate the combined infection was much more virulent and fatal than a pure infection would have been. Moreover, the character of the pneumonia, as given below in the summary of the autopsies, was quite different anatomically, and much more extensive in its distribution than the typical lobar type we are accustomed to associate with pneumococcus pneumonitis.   

It was thought important to make blood cultures in as many of the pneumonia cases as time and equipment would permit, in order to check up on the sputum examinations and to determine accurately the true etiology of the pneumonias. The technique was simply to inoculate 10 to 15 c.c. of blood into 100 to 150 c.c. of plain meat-infusion broth. Plates were not made as a routine but only in specially selected cases. In all 129 blood cultures were made on 111 different patients, 54 of these patients, or 48.7 percent, had one or more positive cultures, while 57, or 51.3 percent, had negative cultures. In other words, practically one-half of the patients had a septicemia associated with their pneumonia.   

The bacteriology of the blood cultures showed Type I, 5 cases, with 2 deaths; Type II, 7 cases, with 7 deaths; Type II (A), 7 cases, with 6 deaths;


Type III, 12 cases, with 12 deaths; Group IV, 19 cases, with 17 deaths; hemolytic streptococcus, 3 cases, with 3 deaths; nonhemolytic streptococcus, 1 case with 1 death; total of 54 cases with 48 deaths, or 88.8 per cent mortality. Two of the 5 Type I cases with septicemia received large doses of specific serum and recovered.   

Only two of the pneumonia cases developed meningitis as a complication. Both showed a Group IV pneumococcus in the spinal fluids and in the blood cultures. One of these patients had such a severe septicemia that it was impossible even to approximate the number of colonies in the plates. Probably in both cases the meningitis was of hematogenous origin. Both patients died soon after the appearance of the meningitis. No cases of influenzal meningitis were encountered.   

There were 29 cases in which thoracentesis revealed the presence of fluid in the pleural cavity. Again, the high percentage of Type III pneumococcus infections was remarkable. The number of sterile effusions (seven) seemed rather large, but corresponded with previous experience that the elaboration of a sterile pleural effusion was not uncommon following pneumonia. Cases of this type got along very well with simple aspiration and practically never required more radical surgical interference. The large number of empyema cases that showed pneumococci in the smears and cultures again indicated that organism as the true etiological factor in the majority of cases of postinfluenzal pneumonia. The influenza bacillus was not found in any of the cultures made from the pleural fluids.   

The type of organism found in the 29 cases of empyema was Type I, 4; Type III, 6; Group IV, 7; undetermined type, 2; hemolytic streptococcus, 3; sterile effusions, 7.   

During the early part of the epidemic 12 autopsies were performed. In eight of the cases there was found a very extensive bronchopneumonia, most marked in the lower lobes, but, as a rule, involving portions of all five lobes. The process was usually so extensive in the lower lobes that the pathology was described as pseudolobar pneumonia. In seven of these eight cases both B. influenzae and pneumococci were demonstrated in smears or cultures from the lung tissue, showing that these patients were suffering from a double infection. Seven of the eight cultures from the heart's blood were positive. Two yielded a Type II pneumococcus, two a Type III pneumococcus, and three a Group IV pneumococcus, and in one of the Group IV cases a pure culture of pneumococcus was also obtained from the spleen. In none of these eight cases was there any pericarditis, endocarditis, or peritonitis. The amount of pleuritis was small considering the extensiveness of the pneumonias, and none had any collection of fluid in either pleural cavity.   

The remaining four autopsies differed essentially from the foregoing and merit separate descriptions. One case showed bronchopneumonia of the left upper lobe, atelectasis of the right lung with a fibrinopurulent pleuritis with effusion (1,000 c.c.) and fibrinopurulent pericarditis with effusion (250 c.c.). Cultures from the affected portions of lung, from the empyema fluid, from the pericardial fluid, and from the heart's blood all yielded pure cultures of hemolytic


streptococcus. This was evidently identical with the hemolytic streptococcus pneumonias which were so prevalent in some of the camps in the past winter, especially following measles.   

Another case proved to be a typical lobar pneumonia involving the right upper and lower lobes with a marked fibrinopurulent pleuritis. Smears and cultures from the affected lung showed many pneumococci but no B. influenzae. The spleen culture yielded pneumococci in pure culture and a Type IIA pneumococcus was obtained from the heart's blood. This case was evidently a straight uncomplicated lobar pneumonia.   

The third case was also a typical lobar pneumonia involving the right upper and middle lobes, with a marked fibrinopurulent pleuritis. In addition, however, there was an acute purulent bronchitis of both lungs. Smears from the right lung showed many Gram-positive diplococci, but the smears from the bronchial exudate of the left lung showed also many Gram-negative bacilli. Cultures from the heart's blood gave a pure culture of Type I pneumococcus. This patient had then both a pneumococcus pneumonia and an influenzal bronchitis.  
The fourth case showed a diffuse bronchopneumonia of both lower lobes, an acute hemorrhagic pleuritis of the right lung, with a slight amount of cloudy, blood-tinged fluid in the right pleural cavity. Cultures from the heart's blood and from the pleural fluid remained sterile. Smears from both lower lobes showed many Gram-negative bacilli and many Gram-positive diplococci. The lung cultures yielded a Gram-positive, bile insoluble, nonhemolytic streptococcus. This patient evidently had a combined influenza and streptococcus infection.  
The postinfluenzal pneumonia may be due to a secondary invasion by any of the ordinary organisms producing pneumonia, but they are apparently more severe, more extensive, and often rapidly fatal. The severity of the secondary infection seems to be augmented by the coexistent influenzal infection. The majority of the pneumonias here were due to a secondary invasion by pneumococci, all types contributing, but Type III and Group IV standing out more conspicuously than Types I and II. The streptococcus, both hemolytic and nonhemolytic varieties, were responsible for a small percentage of the pneumonias. So far as could be determined none of the pneumonias were due to the B. influenzae alone, nor was it possible to recover the B. influenzae from any of the blood cultures or pathological fluids or exudates.   

The pathology and bacteriology of the influenza epidemic at this camp suggest a primary acute infection with the bacillus of Pfeiffer, followed by a high incidence of pneumonia, practically all of which was of the bronchopneumonic variety. The predominant organisms in the latter disease were pneumococci, with a relatively low incidence of the fixed types except Type III, which showed a distinctly higher incidence than in most camps. Streptococcus hemolyticus was relatively rare as a secondary infecting agent. The type of the reaction in the lung appears to have been lobular in the peripheral portions of many of the lungs with confluent spreading pneumonic areas along the bronchial tree, together with lobar or pseudo-lobar types chracterized by incomplete consolidation of whole lobes appearing with relative frequence.


Here, as in other camps, the high incidence indicated in the statistics as primary pneumonia did not confuse the observers, who were quite aware of the preceding acute respiratory infection. It is noteworthy that in this camp, as in some others, the acute respiratory outbreak in the spring of 1918 appeared to have no relation to the reception of new increments of troops into the camp and suggests an increase in the virulence of organisms already present, although with the constant movement in and out of the camp one cannot exclude a virulent organism being carried to it from without and setting up an epidemic condition even in a stablized and seasoned personnel.   

The influenza pandemic occurred shortly after the receipt of a considerable number of unseasoned men, these forming a large proportion of the total personnel of the camp. This undoubtedly accounts in part for the high incidence as shown in Chart I. It is to be noted that the death rate in October was practically the same as in September, yet the morbidity had decreased. In other words, the case fatality rate increased in the latter part of the epidemic.

The chart of incidence and fatality of the acute respiratory diseases at Camp Bowie shows four peaks at which times a mortality above the average for the camp occurred. The first three were due to definite epidemic increases, the last to a high incidence in a smaller personnel present during the demobilization period, representing relatively few cases.   

The material considered in the presentation of the respiratory diseases at this camp consists of 261 necropsy protocols of cases dying of pneumonia, and a study of 2,344 cases of pneumonia and the various complications at the base hospital.e The cases may be grouped into three periods.   

Period 1, from the opening of the base hospital, September 24, 1917, to January 1, 1918: During this period occurred an epidemic of measles, with a large incidence of pneumonia, followed by numerous serious complications and a high death rate.  
Period 2, from January 1, to September 27, 1918: The first three and a half months of this period continued to show a high incidence of pneumonia, though the number and severity of the complications were not as great as in period 1. From April 15 to September 27 was a comparatively quiet period.   

Period 3, from September 28, 1918, to January 1, 1919: This period included a very high incidence of influenza and a high percentage of pneumonia. In contrast to the pneumonia of 1917, there were comparatively few complications, and these were of a less virulent type.   

During period 1 there were 3,624 cases of measles. In addition to the measles there was a widespread infection of the upper respiratory tract throughout the camp. There were 973 cases of pneumonia, and in only 363 of these could there be obtained a history of measles within a month preceding the development of the pneumonia. During the epidemic there may have been in camp

d Sources of information, except as otherwise indicated: (1) Necropsy protocols, Camp Bowie. On file, Army Medical Museum, Washington, D. C. (2) Reports of sick and wounded, 1917 and 1918. On file, Surgeon General's Office.

e The following statements of fact are based, in the main, on: Pneumonia and Some of Its Complications at Camp Bowie, by James C. Greenway, Carl Boettiger, and Howard S. Colwell. Archires of Internal Medicine, Chicago, 1919, xxiv, No. 1, 1-34.

22 CHART II.- The incidence and fatality of the acute respiratory diseases at Camp Bowie


number of cases of abortive measles, unrecognized at the time, in which pneumonia developed later.   

There were 237 deaths among these pneumonia cases, a mortality of 24.4 percent. In 17 percent of the pneumonia cases empyema developed, with a mortality of 32.5 percent.   

The peak was not reached until two weeks after the beginning of the measles epidemic, a much longer time than in the case of the influenza epidemic. There was then a diminution in the number of cases, but following the arrival of 6,000 draft troops, a second peak, almost as great, occurred two weeks after the first. There was a steady decline for 10 days thereafter, when the admissions gradually ceased.   

The pneumonia curve did not follow that of measles so closely as in the later influenza epidemic. The first pneumonia peak followed the first measles peak by 10 days; six days after the second measles peak there was a rise in the number of cases of pneumonia.   

The bacteriology during the period December 6 to 15, 1917, showed, in the sputum examination of 62 cases, pneumococci, Type I, 22; Type II, atypical, 9; Type III, 2, and Group IV, 29. Of the 17 cases in which fluids were examined during this period the hemolytic streptococcus was found in 13, pneumococcus Type II, atypical, in 2, Streptococcus viridans in 2. Lung cultures in 30 cases showed the hemolytic streptococcus in 18. In the period of December 15 to 31, 1917, sputum culture in pneumonia cases of which there were 33 examined, showed pneumococcus, Type I, 9; Type II, 7; Type III, 1; Group IV, 13; pneumococcus Group IV and hemolytic streptococcus, 2; hemolytic streptococcus alone, 1.   

This group of examinations differs considerably from those obtained by the Empyema Commisssion from the questionnaires, charts, and data sent in to them, as it clearly indicates a decided preponderance of the Streptococcus hemolyticus during the fall period.
The first cases coming to necropsy in November, 1917, showed the well known typical lesions of acute lobar pneumonia. Of the first 8 cases 2 showed pleural effusions, 1 serofibrinous, and the other purulent.   

As measles became more prevalent, and as both the incidence and mortality of pneumonia rose, the picture changed. More lobes were involved. Instead of feeling uniformly firm and airless, the affected tissue felt nodular, or "shotty," and distinct crepitation could be elicited between the firm nodules. Fibrinous deposits over the pleural surface were more constant and more extensive. On section, the cut surface of lung presented a mottled appearance, usually a dark, bluish-red background, with many small slightly elevated grayish-red areas, each presenting the opening of a bronchiole, at or near its center. In many cases these areas were confluent. The interlobular septums were distinctly visible, dividing the lobe into a number of irregularly shaped areas. Abscess formation, expecially multiple small abscesses, was occasionally found.   
Effusions into the pleural cavity became increasingly frequent, and were usually purulent. Adhesions between the visceral and parietal pleura occurred, often dividing the pleural cavity into two or more distinct compartments. Perhaps the most frequent site of such adhesion was between the anterior


edge of the lung and the costal margin, the inner surface of the sternum or pericardial sac. Here, elongated pockets were formed, at times divided by transverse bands into two, which contained thick greenish pus and masses of fibrin. Adhesions between the lobes with formation of interlobular abscesses were also found.   

The pericardium was involved in many cases. In some instances only the external surface showed a fibrin deposit of varying thickness with organization, and adhesion to sternum, ribs, and left lung. In other instances, the inner surface was dull, covered with a rough fibrin deposit, and the sac contained a large amount of serofibrinous or purulent fluid, in which hemolytic streptococci were found.   

The miscroscopic picture in these cases was that described as interstitial pneumonia. (See Camp Dodge, p. 62.)   

The pathologic anatomy of the early cases in this period is that of lobar or fibrinous consolidations affecting whole lobes or large parts of lobes, with occasional bronchopneumonic lesions in other lobes. The bacteriology showed pneumococcus Types I and II in about 50 percent of such cases. The streptococci were responsible for the interstitial changes which accompanied some of the lobar consolidations, causing thickening of the septa and giving rise to pleural exudates. There were also peribronchial consolidations such as are found usually in infections with the influenza bacillus. As the epidemic progressed the pathologic anatomy, though still showing some of the lobar types, was preponderantly bronchopneumonic in character. Lesions spreading out from the bronchial tree, and confluent lobular consolidations were seen, together with interstitial lymphangitis and abscess formation. Pneumococcus Types I and II decreased to 33 percent with an increase in streptococci and Group IV pneumococci. Empyema increased also, as did the case fatality. The latter was 3.68 per cent in November, at the height of the epidemic curve. By far the greater part of the pneumonias and empyemas followed measles; there being in November 127 deaths from pneumonia following measles and 45 deaths from primary pneumonia. It is probable that influenza in epidemic proportions preceded many if not most of the cases of primary pneumonia and may well have been a coincident infection with measles in many cases. The case fatality for measles was 4.01 percent, while that for the cases of measles-pneumonia was 39.18 percent.   

During period 2, January and February, 1918, there was a large incidence of pneumonia, the clinical variety showing a greater percentage of lobar pneumonias and differing very little from the variety seen in civil life. Type I pneumonia constituted 26.5 percent of the whole number. Between March 15 and April 15, 1918, 117 cases of pneumonia were admitted to the hospital, 10 of these (8.5 percent) showing Streptococcus hemolyticus in the sputum. The Type I cases dropped to 13.6 percent and Group IV pneumococcus showed 69.2 percent.   

Between March 26 and April 13, 349 cases of an acute respiratory infection were admitted. These were not called influenza because of lack of bacteriologic support for such a diagnosis. Clinically, they closely resembled the condition subsequently called influenza, although this diagnosis is questionable.


During the latter half of June, 1918, a large number of recruits in the detention camp developed severe bronchitis. Sputum from 20 of the more severe cases was submitted to the laboratory for examination. In addition, nasopharyngeal cultures were made from 60 contacts with these cases and from 60 men, chosen at random, throughout the camp.   

The sputum in each case was washed three times in saline and emulsified in broth, then planted in dextrose-blood-broth. After six to eight hours' incubation, human blood-agar plates were streaked from these cultures and were studied after 12 to 18 hours' incubation, and again after 48 hours' incubation. Pneumococcus colonies were fished into dextrose-blood-broth, and pure cultures were submitted to the test for bile solubility and agglutination with type serums. Colonies showing a wide area of hemolysis and having morphologic characteristics were considered hemolytic streptococcus.   
In obtaining nasopharyngeal cultures, the nasopharynx was swabbed as for meningococcus carrier detection. Swabs were placed immediately into plain broth tubes, and from these, human blood-agar plates were streaked and incubated for 24 to 36 hours. Plates then were studied for types of colonies. Organisms such as staphylococci and Micrococcus catarrhalis were not recorded.   

The result of this examination showed hemolytic streptococci in 100 percent of the 20 cases of more severe type, in 3 of which there were also pneumococcus Type I, and in 15, pneumococcus Group IV. Of the 60 nasopharyngeal cultures, 3 showed hemolytic streptococcus, 7 pneumococcus type undetermined, 8 pneumococcus and hemolytic streptococcus, and 20 neither pneumococci nor streptococci, other organisms than these not being recorded.   

During the remainder of period 2, especially after the 36th Division left in July, there were only 43 cases of pneumonia with 2 deaths. Nine of the 43 showed the hemolytic streptococcus predominating in the sputum, and 2 were a mixed infection of hemolytic streptococcus and Group IV pneumococcus.   

Of the 538 pneumonia patients, only 48 gave a history of an antecedent infection within a month.  
The total mortality for this period was 7.4 percent, there being 40 fatalities among the 538 cases. Fifty-two patients (9.7 per cent) developed empyema and provided 37.5 per cent of the total deaths.   

It is significant that in all the lobar pneumonia cases pneumococci were found in the sputum, and 10 of the 18 fatal cases reacted to specific type serums I and II. Hemolytic streptococci were still an important factor, and in three of nine purulent pleurisies this organism was present in addition to the pneumococcus. During the month of April there occurred four cases of interstitial bronchopneumonia caused by the hemolytic streptococcus. These cases occurred at the time of the small epidemic of acute respiratory infection previously described. Of 25 post-mortem examinations made during this period, lobar pneumonia was found in 19 cases; bronchopneumonia, of the interstitial type, was found in 6 cases. The incidence of purulent pleurisy was high (50 percent). This may be explained by the fact that five of the six bronchopneumonia cases showed this complication, and hemolytic streptococci were found as complicating organisms in one-third of the cases following pneumococcus infection.


After the epidemic wave of October, 1917, to January, 1918, the fixed type pneumococci again preponderated in the pneumonias until March-April, when the streptococcus hemolyticus invasion increased and the type of pneumonia present, lobar, in considerable proportions at first, became almost wholly of the bronchopneumonic, lobular, and interstitial varieties. A definite influenza outbreak occurred in April and, though the streptococcus appears to have been prevalent, the case fatality was only 0.45 per cent of the total respiratory disease. Recruits arriving at the camp picked up the prevailing infection with the streptococcus and other organisms. The bacteriological examinations in June showed the flora present in the interepidemic period and preceding the wave of the pandemic.   

The development of the epidemic, period 3, on September 17, began with the admission of cases of a not uncommon type of upper respiratory tract infection, numbering about five daily, before the semblance of an epidemic was evident. On September 26, 24 cases of influenza were admitted to hospital. During the epidemic, which continued to November 15, 1918, 3,876 cases of this infection were admitted to hospital; 634 cases were held in the observation wards of the detention camp.   

Study of the blood counts in uncomplicated influenza showed that during the first five weeks, representing the period of highest incidence, the largest number of counts was below 10,000; the polynuclear percentage was below 70, and the lymphocyte percentage was over 25. As the severity of the epidemic declined, the total white counts were higher, and the polynuclear percentages were correspondingly increased. The same transition was apparent in the pneumonia following influenza, although the percentage of high counts was greater than in the uncomplicated cases. Leucopenia (counts under 5,000) was not a common finding, though absence of leucocytosis was the rule.
This change in the blood findings was synchronous with a somewhat changed clinical picture. The epidemic reached its peak in numbers in seven days, and at the end of the second week the severity of the infection was decreasing, as shown by a less profound prostration, diminished general malaise, less frequent and milder toxic erythema, and more rapid convalescence.   

The first peak of the influenza preceded that of the pneumonia by five days. The second peak, occuring three weeks later, coincident with the arrival of the first increment of drafted troops from Missouri, preceded the second peak of the pneumonia by seven days. There were received at the detention camp during the course of the epidemic 4,108 drafted white men and 2,360 drafted colored men.   

Among the white men, 252 per 1,000 men who were exposed developed the infection, 114 per 1,000 developed pneumonia. Not every pneumonia patient gave a history of an antecedent influenza infection. Among the 795 cases of pneumonia, a history of influenza within the preceding month was obtained in 728.   

No deaths were attributable to uncomplicated influenza. The total deaths in the 833 cases of pneumonia, numbered 156, a mortality of 18.7 per cent. Thirty-six men (4.3 per cent) developed empyema and furnished 11.5 per cent of the total deaths.


In spite of the higher morbidity rate for influenza and pneumonia among colored troops, the mortality percentage (17.1 percent) among them was somewhat less than that of the white troops (20.8 percent).   

It would appear at first that the highest mortality was reached quickly after the onset of the epidemic, the virulence of the infection decreasing thereafter and the mortality diminishing from 21.2 and 19.1 percent to 12 and 9.1 percent, respectively, for semimonthly periods. That this may have been partly true is evidenced by the fact that the patients admitted after the first two weeks did not appear to be so severely ill. That the virulence of the infecting organisms persisted is shown by the incidence and mortality of pneumonia among the Missouri drafted men, who arrived three weeks after the peak of the epidemic was reached. These men arrived in camp from October 22 to October 25. They were overtaken by influenza October 23 to 31 and developed pneumonia October 28 to November 4. Among them there developed 199 cases of pneumonia with 39 deaths, a mortality (19.6 percent) almost equal to that occurring at the height of the epidemic. The lower total mortality for the first half of November was due to the inclusion of 91 scattered cases occurring among troops that had been in camp longer. Among these cases there were only 11 deaths, a mortality of 12 percent.  
On September 25, 1918, routine nasopharyngeal swabs were made in all cases admitted with the diagnosis of influenza. In all, 75 cases were examined satisfactorily.   

Swabs were made as for meningococcus carrier detection. These were streaked first on human blood-agar plates, and afterwards immersed in glucose-blood-broth. The latter cultures were incubated about six hours and human blood-agar plates then were streaked from them. The identification of streptococci and pneumococci was in accordance with "Standard Method for U. S. Army, " except that press of other work prevented identification of types of pneumococci. Plates were incubated for 18 to 24 hours. Influenza bacilli when found in quantity were present in characteristic growth in this time. The nasopharyngeal cultures showed hemolytic streptococci preponderating in 45 percent of the cultures, pneumococci, undetermined, in 35 per cent, influenza bacilli present in 9 per cent, nonhemolytic streptococcus colonies predominating in 10 per cent. Fixed type or Group IV pneumococci predominated throughout the epidemic, with a sharply rising incidence of streptococcus toward the latter part, followed by a decline and a sharp rise in Group IV incidence. The bacteriology of the lobar pneumonia by lung cultures and of the bronchopneumonia showed a preponderance of pneumococci, type undetermined, and Group IV, 2 cases of Type I, 2 cases of Type III in 68 examinations, while the hemolytic streptococcus was found alone in 2 and associated in 12. The influenza bacillus was found in 16 percent of the cases of bronchopneumonia and 25 percent of the cases of lobar pneumonia. The streptococcus predominated in the pneumonias of interstitial type, being found alone in 10 of 33 cases and in 64 percent of all cases, while the influenza bacillus was present in 2, or 9 percent of the cases. Empyema was most frequent in the interstitial type as compared with the other two types. The bacteriology of the heart's blood at necropsy showed pneumococci, type undetermined, in 19, Group IV in


17 Type I in 1, Type II in 2, Type III in 1, hemolytic streptococcus in 11, and nonhemolytic streptococcus in 11; total cases examined numbered 81, in 19 of which there were no growths.   Three distinct types of cases were found: (1) lobar pneumonia; (2) bronchopneumonia of interstitial type; and (3) bronchopneumonia of a lobular or confluent type.   

In cases of the third type, the lungs were large, moderately firm, all more or less nodular. Frequently, by careful palpation, crepitating areas could be felt, even in a lobe which was apparently completely consolidated. The pleural covering was usually smooth and transparent, many cases showing scattered small areas of subpleural hemorrhage. Fine fibrinous deposits were found in some cases very early in the course of the pneumonia. In cases which showed emphysema of the mediastinal tissues there was usually an emphysema just beneath the pleura which extended mainly along the interlobular septa.   

Section of the lung was followed by the outpouring of a large amount of dark, bloody fluid. The cut surface was dark, bluish-red, mottled with lighter areas, which varied in color from a deep brownish-red to a light grayish-red, or grayish-yellow, and varied in size from one or two centimeters in diameter up. It was very common to find areas of different color and consistency in the same lobe. In the lower lobes these areas were largest and most confluent, so that at times it was difficult to say whether or not the consolidation was homogeneous. In such cases, however, the upper lobes always showed distinct patchy consolidation.   

The characteristic features of the microscopic pictures were the intense edema and congestion of all interstitial tissues, the marked predominance of bloody exudate in the alveoli in most cases, and the irregular distribution of the lesion, not strictly peribronchial nor diffuse, but showing areas at different stages of development in the same section.  
Empyema as a complication occurred in 249 (10.6 percent) of the 2,344 cases of pneumonia. This incidence was a steadily decreasing one for the three periods: 16.5, 9.5, and 4.3 percent, respectively. Period 1 furnished 41.5 percent of the total pneumonias, and 64.6 percent of the total cases of empyema, whereas period 3 furnished 35.5 percent of the total pneumonias, and only 14.4 percent of the total empyemas.   

Metastatic abscess, single or multiple, occurred in 17 cases of pneumonia. Of these 17 cases, 4 accompanied empyema; 5 came to necropsy, 3 not being diagnosed before death. One of these undiagnosed cases was a small abscess in the first intercostal muscle; a second case was a perirenal abscess. The most interesting cases of this group were five cases in which a painless tumor was noted just above the symphysis pubis, not red and not tender to pressure. The first of these cases was not recognized until there appeared a rounded tumefaction exactly in the midline, extending about 2 inches above the symphysis pubis. The tumor was thought to be a distended bladder, but on catheterization a very small amount of urine was obtained; 100 c. c. of salt solution was introduced into the bladder, and the same amount was withdrawn. The tumor was still present. The possibility of a diverticulum of the bladder was considered. There was no increase in temperature, and no muscular


rigidity. During the next six days there was a slight, but constant, increase in the size of the tumor, the patient meanwhile voiding a normal amount of urine. On the sixth day fluctuation was obtained. An incision was made and about 150 c.c. of thick pus was found between the right rectus muscle and its posterior sheath. Culture showed a pneumococcus, type undertermined. The other cases presented practically the same picture, except that in three the tumor was slightly asymmetrical.   
Generalized subcutaneous emphysema, as a complication of pneumonia, occurred in 8 cases, 7 white men and 1 colored man; 3 of these patients recovered.   

The onset of the emphysema appeared to be definitely associated with an unusually severe paroxysm of coughing in one case. Hoarseness and increased dyspnea were the only symptoms apparently referable to this complication.   

The tissues of the neck, posteriorly especially, but also in the supraclavicular region, were involved in all the 8 cases; the interscapular region was involved in 4; the axillme and pectoral region in 3; the face in 3; the flanks and thighs in 1.
The three patients who recovered had a prolonged convalescence. Each one had a persistent hoarseness. Laryngeal examination showed only an acute laryngitis, which disappeared slowly. One patient has since been found to have pulmonary tuberculosis.

Anaerobic ante-mortem blood cultures were sterile in two cases; in the third case they showed pneumococcus, type undetermined.   

Of the post-mortem cases all showed emphysema of the mediastinal connective tissue, in addition to the evident subcutaneous emphysema noted during life. Petechial subpleural hemorrhages were evident in two cases. Free air beneath the visceral pleura, especially along the interlobular spaces and lung fissures, was seen in two cases. One case showed confluent patches of consolidation having undergone softening and produced abscess cavities. Microscopically small areas of hemorrhage were frequent in these cases as in other pneumonia cases of the same period.   
In addition there were 10 cases of emphysema of mediastinal tissue diagnosed only at necropsy.   
Cultures of the mediastinal tissue showing marked emphysema in 8 cases, which gave no growths in 3; pneumococcus Group IV in 3; streptococcus, nonhemnolytic, in 1; streptococcus, hemolytic, in 1. In each case the organism present was found also in the blood culture.   

Certain cases, especially of lobar pneumonia, offered no difficulty in diagnosis. In the secondary pneumonias, those following measles in 1917, and especially those following influenza in 1918, it was the rule rather than the exception for the signs to be so irregular as to offer the greatest difficulty in diagnosis, not only as to the evidence of consolidation, but especially as to the clinical variety. The sputums from these cases practically all showed streptococcus, pneumococcus Group IV, or a mixed infection.   

Considerable difficulty was experienced in making correct diagnoses as to the type of pneumonia. In October and November, 1917, 50 percent of the cases of bronchopneumonia were wrongly diagnosed as lobar, 45 percent were wrongly diagnosed in September and October, 1918. Even post mortem the


differentiation was frequently puzzling and the data as reported on the sick and wounded records are stated to be distinctly unreliable.
A study of the protocols shows that most of the lungs were affected by a mixture of types of pneumonia not all of which appeared to be of the same duration. Early symptoms therefore were the result of one lesion which may or may not have preponderated throughout the clinical course and at the necropsy.   

Positive blood cultures in this camp, as in others, indicated an unfavorable prognosis. The mortality by type of organism in the pneumonia cases showed, for Type I, 4.9 percent; Type II, 20.6 percent; Type III, 13.3 percent; Group IV, 10.4 percent; Group IV and hemolytic streptococcus, 24 percent; nonhemolytic streptococcus, 100 percent; hemolytic streptococcus alone, 13.3 percent. Mortality of all types of organisms was distinctly increased during the influenza epidemic, with the exception of Type III which showed an apparent decrease.   

Experience at this hospital showed that to limit the investigation of sputum to a determination of the type of pneumococcus present, does not always give a reliable guide as to the nature of the existing infection.   

Three waves of influenza or influenza like disease occurred at this camp during the World War. The variations in the proportions of the types of lesions in the lung governed the varying physical signs and clinical course in those affected and were dependent on the type of organism causing the disease. Those of the personnel who had passed through an epidemic showed a decreased incidence to influenza or influenza like disease and to the complicating pneumonias and also a lower case fatality rate than did the unseasoned recruits. Lobar consolidations appear to have been caused by pneumococci with a large proportion of the fixed Types I and II. Streptococcus not only produced characteristic lesions alone but also infected lungs already the seat of pneumococcus lesions, producing interstitial lesions and empyemas. The bronchopneumonias lobular, interstitial, and spreading confluent types-were associated with pneumococcus Group IV and streptococcus, hemolytic and nonhemolytic, the latter producing most of the interstitial pneumonic reactions.   
While the bacteriological examinations at this camp were excellent it is doubtful if the total incidence of B. influenzae was determined. The fact that the respiratory lesions spread downward along the air passages as stated by the clinicians and as indicated in the necropsy descriptions strongly suggests an infection with this organism. The frequency of multiple infections with different bacteria is well brought out in the necropsy protocols at the Army Medical Museum and in the study made by the hospital staff.   

Chart II indicates that an increased mortality from respiratory disease occurred just preceding the influenzal waves of the spring and fall of 1918. The case fatality rate for all respiratory diseases was practically the same for the epidemic of 1917 and the pandemic incidence at this camp in 1918, 3.68 percent in November, 1917, and 3.86 percent in October, 1918. How much these figures would be changed had all of the light cases been recorded can not be determined. With the higher incidence of October, 1918, it might well reduce the figure for that month.


The epidemic of respiratory diseases at Camp Cody in the fall of 1917, aside from measles, was due to influenza and common respiratory diseases. During the beginning of this outbreak, or in the month of November, the case fatality rate for all respiratory diseases was 0.95 percent. With the continuation of the epidemic, the streptococcus appeared in considerable numbers and the case fatality rate increased to 1.71 percent in December. The deaths during these two months were charged very largely to the primary pneumonias, of which lobar pneumonia exceeded, so far as can be judged from the records of sick and wounded. The few autopsy protocols and specimens available from this camp do not include any for this period, but the records of the Empyema Commission indicate a considerable incidence of streptococci. These organisms may have been secondary invaders of cases primarily affected by pneumonia of lobar type, but it is much more probable that the pneumonic lesions were of the usual type seen in other camps in the majority of which the typical fibrinous consolidation of cropous pneumonia was rarely seen. Influenza cases increased in March, 1918, without increase in case fatality, as is indicated by Chart III. The sharp rise in June followed the receipt of between four and five thousand men, and the streptococcus again was prevalent at this time, the case fatality rate being 1.95 percent for this month. The case fatality dropped in July to 0.46 per cent, but rose sharply in August to 3.73 per cent preceding any apparent clinical evidence of the beginning of an epidemic. In September the case fatality rate dropped to 2.83 percent, there being the same number of deaths in September as in August, namely 6, while the incidence increased from 161 cases in August to 212 in September, apparently the beginning of the influenza pandemic. The case fatality jumped to 6 percent in October and increased to 12.75 percent in November, with approximately a third as many cases occurring as in the previous month. The streptococcus had a large part in producing this increase in the case fatality rate.g   
On September 24, 1918, there was admitted to one of the general medical wards of the base hospital at Camp Cody, a soldier who had just come from Camp Dix, N. J., with prisoners. He gave a history of having been sick for three days, on the train, with headache, rhinitis, general aching, soreness in the chest, and cough. In addition, he presented a well-marked pharyngitis. Within 24 hours from admission, certain unusual features were present. The pulse was slow in proportion to the high temperature. The respiratory rate was greater than physical signs warranted, and the leucocyte count was 4,800. Repeated physical examinations on that day finally revealed a few crepitant rales near the angle of the right scapula. In the afternoon of this second day, cyanosis of the finger tips became noticeable, and an ashy color of the face appeared. The condition in the lung spread rapidly to the left lower lobe posteriorly, and on October 1 the man died. This was the introduction of this camp to the acute respiratory disease which was then prevalent in the northern and eastern camps.

f Sources of information, except as otherwise indicated: (1) Medical reports to the Surgeon General, 1917, 1918, and 1919. (2) Empyema records, on file, Surgeon General's Office. (3) Necropsy protocols, on file, Army Medical Museun.  g The following statements of fact are based, in the main, on: The Epidemic of Respiratory Infection at Camp Cody, N. Mex., by Frederick H. Lamb and Edward B. Brannin. Journal of the American Medical Association, Chicago, 1919, lxxii, No. 15, 1056-62.

32 CHART III.- The incidence and fatality of the acute respiratory diseases at Camp Cody


The epidemic here may be said to have begun, October 1, 1918, with the admission of nine cases which were diagnosed influenza, and, from the standpoint of this study, it closed, December 1, 1918.   
Laboratory investigations were made during the first 10 days of the epidemic in 80 typical cases of influenza. At the time of examination, none of these patients presented any physical signs of pneumonia or other complication. Nasopharyngeal and sputum cultures, blood cultures, blood counts, and urine examinations were made in practically all of these cases.   

Assuming that droplet infection was the most important factor in the dissemination of the disease, the bacteria of the nasopharynx and sputum were studied in detail. Nasopharyngeal swabs, obtained by means of West tubes, and specimens of sputum collected in sterile Petri dishes, were inoculated into blood agar and serum boullion. Considerable pains were taken to identify colonies by subcultures in differentiating media. In order to gain some idea of the virulence of the organisms, 58 white mice were given intraperitoneal injections of sputum.   

It was found that there were five predominating pathogenic organisms: Pneumococcus, B. influenzae, M. catarrhalis, staphylococcus and streptococcus, both hemolytic and nonhemolytic. Intraperitoneal inoculations of the pneumococcus, staphylococcus and both forms of streptococci were fatal to white mice. Injections of cultures of B. influenzae and M. catarrhalis produced only toxic symptoms in the mouse.   

In 80 cases investigated, the organisms recovered from both the throat and sputum cultures showed pneumococcus present in 90 percent; B. influenzae in 46 per cent; M. catarrhalis in 84 percent; hemolytic streptococcus in 14 percent; nonhemolytic streptococcus in 26 percent, and staphylococcus in 20 percent. Colonies of both B. influenzae and pneumococcus were found in 41 percent of the plates. A few plates showed colonies of all these organisms. In many cases it was difficult to tell which was the predominating organisms, and such a classification was given up as being unsatisfactory. Of the 58 white mice that received intraperitoneal injections in 1 c.c. doses of ground sputum, 9, or 18 percent, died. From seven of these the pneumococcus alone was recovered in the heart blood; from two, the pneumococcus and B. influenzae were recovered. The foregoing bacteriologic observations indicated that B. influenzae was not the specific organism.
Fifty-four blood cultures were made in this series of 80 cases, all of which remained sterile.   
The average of 80 leucocyte counts was 6,780. In 54 cases, or 67 percent, the count was less than 8,000. The highest count of the series was 22,600, and the lowest 3,300. The differential count was not remarkable.   

Sixteen specimens of urine, or 20 percent out of 78, showed a trace of albumin. Two cases showed considerably more than a trace, with a few granular and hyaline casts.   

In connection with the experimental work, it is interesting to note that very soon after the epidemic of influenza reached this camp, laboratory guinea pigs, housed in a small building beside the laboratory, began to die. At first it was thought to be the result of food poisoning, but a necropsy on a dead pig revealed


unmistakable signs of pneumonia. The mucosse of the trachea and bronchi were deeply injected and covered with a glairy, serofibrinous exudate. Cultures and smears from this exudate and pieces of lung tissue revealed pneumococci in large numbers, a few streptococci and small Gram-negative bacilli, identical in every way with B. influenzae. These observations were confirmed in subsequent necropsies. The guinea pigs were sick from two to four days before death. During this time, the sick animal, trembling from chills, with hair ruffled, sat huddled up in a corner of the pen, moving about only to eat. This it would do until shortly before death. The respirations were rapid and wheezing; the characteristic shrill whistle became scarcely audible. The animal was apparently in a stupor which gradually deepened until death supervened. Altogether, it presented the picture of a profound intoxication. Just before death the animal would fall on one side, rise a time or two, then make a few feeble efforts to do so again. Within 15 or 30 minutes it would die after several rather weak, clonic, convulsive movements.   

All of the pigs in a batch of 30 died within three weeks. Fifty more were received from El Paso, Tex., just about this time, four of which were dead on arrival. The others were apparently healthy. These were placed in a different room of the animal house, which had not been in use for several months. It was thoroughly cleaned and supplied with fresh hay, which was changed daily. A heater was placed in the house, and every effort was made to keep this new batch clean and healthful. No deaths occurred for about two weeks. One morning one pig was found dead, and within three weeks all of this second group had died just as the first had done. Segregating the sick animals did not save the others.   

Pneumonia occurred in 624 Army cases. In type, with few exceptions, it was of a lobular variety. Clinically it was considered a hemorrhagic atypical bronchopneumonia. The massive consolidations were thought to be the result of the confluency of contagious inflamed areas.   

Blood cultures in 248 cases of influenza-pneumonia showed 4 positive results; hemolytic streptococcus in 1 case and pneumococcus in 3, of which 1 was Type II, 1, Type II atypical, and 1, Group IV. These findings contrast with those of the blood cultures made in the spring and summer preceding, where 26 percent of 92 and 18 percent of 180 cases were positive. Typing of the sputa with white mice and Avery's blood broth was made in about 50 percent. Typing did not have the satisfactory results recorded during the year previous to the pandemic period. The organisms found were pneumococcus Type I, 2.6 percent, Type II, 2.01 percent, Type II, atypical, 6.89 percent, Type III, 4.02 percent, Group IV, 46.55 percent. In the examination of 91 cases hemolytic streptococcus was found 37 times and nonhemolytic streptococcus 54 times. The organism recorded was considered to predominate in the cultures, other organisms also being present.   
The comparative incidence of the five most frequently found organisms in the pneumonia cases were pneumococcus in 89.1 percent, streptococcus in 63.3 percent, B. influenzae in 23 percent, M. catarrhalis in 6 percent, and staphylococcus in 8.6 percent.


Taken as a whole, the patients who died were, if anything, somewhat better nourished and more robust than the average individual. The rapidity with which death took place precluded much external change. Post-mortem lividity usually was marked; the finger nails and mucose were very cyanotic. In about one-half of the cases a foamy, blood-stained liquid ran from the nose and mouth when the head was lowered.   

In all except one of the 16 cases examined at necropsy, the post-mortem picture was strikingly uniform. This exceptional case presented all the characteristics of a true, dry, lobar pneumonia.   
The mucosm of the trachea and bronchi were intensely red, slightly edematous, and covered with a red, mucopurulent exudate. In many instances the smaller bronchi were filled with a red foam. The peribronchial lymph nodes were twice or thrice normal size, very red, soft, and the cut surface granular.   

No changes were noted in the mediastinal space.   

The pathologic changes in the lungs were certainly not those one is accustomed to see in the ordinary lobar pneumonia, nor were they typical of bronchopneumonia. In fact, the terminal hypostatic pneumonia, sometimes seen in the aged, compares more nearly, in gross characteristics, than does either of the other types.   

In the 15 cases designated as atypical, hemorrhagic bronchopneumonia, the pneumonic process was bilateral. In nine cases it was of nearly equal extent on the two sides; in six it was decidedly more pronounced on one side than on the other. In three cases there were extensive, dense, pleural adhesions; in four, there were recent, friable adhesions; in eight, the lungs were free. In two cases, there were 500 c.c. and 700 c.c. quantities, respectively, of a dark red, serosanguineous exudate. Both parietal and visceral pleurae were deeply injected. The lungs were voluminous and heavy. As a rule the posterior half or three-fourths of the lung was a dark, purplish red, the remaining anterior portion, gray or pink. That this appearance was due to the filling of the posterior portion with dark red blood could be clearly demonstrated by an anteroposterior section. The posterior portion felt uniformly firm, yet somewhat resilient. The anterior part, especially of the middle and upper lobes, was elastic, soft, and air containing; here sometimes a discrete, irregular, consolidated area could be felt.   

On section, a profuse bloody exudate welled from the cut surface of the consolidated portion. With slight pressure, 300 c.c. of this liquid could be drained from a single longitudinal section. That the firmest portions of the lung contained air could be demonstrated by the presence of numerous air bubbles in the exudate. Small sections from the more dense areas would barely sink in water; other sections floated. After the surface of a section had been squeezed and washed, the polygonal boundaries of the lobules could be plainly seen. A section through the anterior, air containing portion, which was rather sharply demarcated from the posterior consolidated portion, revealed a comparatively dry cut surface. Twelve out of 30 lungs presented peribronchial areas of consolidation from 1 to 4 cm. in diameter, in this air containing portion. The cut surfaces of these isolated areas bulged slightly, and were


dark gray but not hemorrhagic. A mucopurulent exudate could be expressed from the bronchioles involved. These areas were typically those of bronchopneumonia. The marked hemorrhagic, edematous element was entirely lacking in this upper and anterior half of the lung; in fact, the gravitation of the blood and exudate to the posterior of the lung was striking.
Gross changes in the heart or pericardium were not encountered. No evidences of dilatation on either side were found; the muscle was firm; the ventricles, usually contracted, contained a small amount of dark, semiclotted blood. The "chicken-fat" clots, often observed in pneumonia and slow deaths, were not found. Only the changes usually resulting from the toxemia of a severe infection were noted in the other viscera. No marked evidences of nephritis were found. The average weight of the spleen was 240 gm. It was usually of firm consistency; the cut surface moist and purplish red. A small amount of red, gruel-like pulp could be scraped from the surface. The bone marrow was pale, and quite noticeably dry, an observation in keeping with the leucopenia.   

Bacteriologic cultures were made from the lungs and the heart's blood in 4 cases.   
Of the complications, the most frequent was that of pleuritis with exudate. In the group of fatal cases, a serofibrinous effusion of sufficient quantity to merit attention was recorded in 20 instances. Seven of these became purulent. In the nonfatal group it was more common, and resulted in 30 empyemas. The bacteriology of the fluids examined showed the streptococcus to be present alone in 36 percent, the pneumococcus alone in 36 percent, the streptococcus and pneumococcus associated in 20 percent, and the staphylococcus in 8 percent. A striking feature was the frequency with which the accumulation, probably beginning in a fissure, would press the lobes apart, giving a V-shaped area of fluid as seen with the roentgen ray, with the point of the V at the hilum of the lung and the open part at the chest wall. Physical examination in these cases naturally showed an area of compressed lung both above and below the fluid.   

Otitis media was infrequent, although next in incidence, to the pluerisy. Phlebitis of the left leg occurred in three instances. Furunculosis and a tendency to small abscess formation were noted in a few cases. One case of cerebral thrombosis was recorded. Pulmonary edema, as a terminal condition, often was seen. Pericarditis was not found, and but one instance of acute endocarditis was observed. Meningitis with pneumococci in the fluid was recorded seven times.   

Relapses or recurrences were unusual. Remissions in the severity of the process with subsequent intensification of symptoms, as early involved lung areas improved and new areas became infected, were seldom observed. Two cases of subcutaneous emphysema were recorded; both patients recovered. Spontaneous pneumothorax occurred in two instances. The picture of mild neurocirculatory asthenia was seen many times in convalescents. Reactivated, chronic tuberculosis was an unusual finding in spite of the fact that many patients had never had their induction examination. Very few mental or nervous sequelae were observed. It was not unusual to find, long after recovery, the per-


sistence of the physical signs of fluid in the bases, in patients in whom repeated aspiration had failed to establish its presence. This was thought to be due more often to a fibrinous pleural exudate than to an unresolved pneumonic process. Crepitant and suberepitant r?les in the bases were found in some instances to persist for a period of four or five weeks from the onset of normal temperature.   

This camp showed the usual mobilization epidemic of measles and acute respiratory diseases, with the medical officers frequently calling acute respiratory affections "influenza." There was no apparent epidemic in the spring. Although the rise in June may possibly be considered such, it is more likely an increase due to a considerable number of raw troops being received, and was accompanied by the sharp rise in measles. The only bacteriological data available for the cases of the fall and spring are those of the Empyema Commission which, as previously noted, show that there was a very definite high incidence of streptococcus. The indications are that the pneumonias were less hemorrhagic in character in the fall of 1917 than in October and November of 1918. During the influenza epidemic, the presence of the streptococcus was an important factor in the pulmonary lesions. In the month of October the ratio per 1,000 for empyema was over 3, the rate for December, 1917, being 1.7, and in January, 1918, 1.2. The nonhemolytic streptococcus apparently predominated over the hemolytic during the influenza period.   

There are 12 protocols with specimens from 7 cases from the influenza epidemic at Camp Cody in the accessions of the Army Medical Museum. The organisms described bacteriologically can be found in the sections, and in addition in well-fixed specimens minute Gram-negative bacteria are abundant along the walls of the smaller air passages and in the infundibula, particularly in the hemorrhagic lesions and in the periphery of the lung. Pathology at this camp differs in no way from that of several other camps where the streptococcus predominated, giving a high proportion of interstitial and hemorrhagic types of pneumonia, though all of the ordinary types are represented. The camp peculiarity is the lack of a wave of epidemic influenza in the spring months, while no camp shows better than this one the increased case fatality during the latter part of the influenza pandemic.


Coincident with mobilization at Camp Custer, acute respiratory diseases occurred in epidemic proportions, with a case fatality rate for all respiratory diseases of 0.32 percent in December, 1917, rising to 1.15 percent in January, 1918. In addition to measles, influenza was diagnosed in the cases during this period, though the deaths therefrom were attributed, in the official statistics, largely to primary pneumonia. The presence of the streptococcus among respiratory cases became more noticeable during the month of January, contributing to the increase in the case fatality rate. Following the recession of all respiratory diseases in February, there was a sharp rise extending over March and April, the principal increase being due to cases of influenza. The total case fatality rate for respiratory diseases was 1.67 percent in March and 2 per-


Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1917, 1918, and 1919.

38 CHART IV.- The incidence and fatality or the acute respiratory diseases at Camp Custer


cent in April. During these two months the streptococcus was commonly present, as is shown by the records of the Empyema Commission. This wave of respiratory diseases was definitely influenzal in nature, though the deaths were attributed to or diagnosed as primary pneumonia. Here, as in other camps, while the cases may have been admitted with influenza or acute respiratory disease, the diagnosis was frequently changed to pneumonia and the deaths were so recorded.   

Prior to the influenza epidemic, and in anticipation of its occurrence, the personnel of the camp made a study of the prevalent mouth flora of 357 healthy individuals.i Smears and cultures on plain and blood-agar plates were made from the nose and nasopharynx. The usual mouth organisms were recovered. Seventy-five percent of the cultures showed hemolytic streptococcus; in only five cases (a little more than 1 percent) was the influenza bacillus identified.   

After the epidemic had begun, cultures were made from the throat in 366 cases, which showed no physical signs of pneumonia, plain blood-agar plates being used. The organisms found were hemolytic streptococci, 34 percent; nonhemolytic streptococci, 33 percent; pneumococci, 12 percent; influenza bacilli, 8 percent. One hundred and forty blood cultures made on these patients remained sterile. Blood counts showed leucopenia as the most impressive feature, 70 percent of the counts being below 8,000. Twenty percent of the urinalyses showed albumin, in the majority of cases in small amount; casts occurred in 4.7 percent.   

In the cases with pneumonia the hemolytic streptococcus was the most common organism found in the blood, but only 2 percent of 510 blood cultures in cases of pneumonia showed organisms, whereas in the previous winter, 20 percent of the streptococcus pneumonias were positive. There appeared to be an increase in the number of positive cultures during the latter part of the epidemic.   

Typing the sputum was found much less satisfactory than had been the experience during the year preceding, and the reactions with immune sera were less clear-cut. The organisms isolated from the sputum of pneumonia patients showed 25.8 percent pneumococci, 20 percent of these being due to Group IV. Streptococci were found in 60.9 percent of the cases, 43.9 per cent nonhemolytic, 17 per cent hemolytic. The influenza bacillus was found in 5.2 percent. In only eight cases it was the organism found alone.   

In the blood counts 67.8 per cent showed 8,000 or fewer leucocytes per cubic millimeter. Albumin was present in the urine of 40 percent of these cases, and casts were found in 22 percent of the urines examined.   

The incidence of organisms as found at necropsy in the lungs of pneumonia patients (280 cases) were Type I, 8; Type II, 34; Type III, 18; Group IV, 18; streptococcus, hemolytic, 76; nonhemolytic, 66; influenza bacillus, 8. The incidence of organisms as found at necropsy in the heart's blood of pneumonia patients (280 cases) were Type I, 6; Type II, 33; Type III, 16; Group IV, 22; streptococcus, hemolytic, 62; nonhemolytic, 48; influenza bacillus, 3.

i The following statements of factarebasedon: A Recent Epidemic of Acute Respiratory Infection at Camp Custer, Mich., by Wyndham B. Blanton and Ernest E. Irons. Journal of the American Medical Association, Chicago, 1918, lxxi. No. 24,1988.


Without exception the deaths from this respiratory epidemlic were due to secondary pneumonia. In no instance did a case come to necropsy in which death occurred from influenzal infection alone. The bodies were those of well nourished young men, the termination having been reached too quickly for gross external changes. Many of the bodies showed a moderate degree of corpulence. Associated chronic lesions occurred in sufficient number to warrant the belief that in a fair proportion of fatal cases the patients were seriously handicapped beforehand in the battle with the infection.   

For the most part there was a striking similarity in the appearance of the respiratory organs in the 123 cases of bronchopneumonia coming to necropsy. The mucous membrane of the trachea and bronchi appeared intensely red, swollen, and covered with a mucopurulent exudate. The bronchial glands were enlarged, soft, and reddened on section. In the majority of cases the lungs were voluminous and heavy. As a general rule the posterior lobes varied from the remaining lobes, and presented a surface, dark purplish red, sometimes smooth, sometimes roughened and dull from collections of fibrin. They felt firm throughout, and only occasionally could areas of greater density be made out.   

The remaining lobes showed a greater proportion of air-containing tissue. They usually presented pale, nonresilient emphysematous patches interspersed with dark red, firm, and slightly depressed areas, which were to be felt as irregular consolidations extending to various depths into the lung substance. Section of the lung throughout the lower lobes usually displayed an exceedingly moist cut surface, the slightest pressure forcing to the surface quantities of blood-tinged fluid. Sometimes this existed to such a degree as to obscure the underlying process. Again, the confluence of separate patches of consolidation was sometimes so complete as to be confusing, but careful study usually made evident the bronchial distribution of the process. As a general occurrence, however, patchy areas of consolidation clustered about the bronchi were easily made out. Mucopurulent plugs filled the bronchi and bronchioles. Section of the remaining lobes showed a much drier lesion. Here islands of dark red, often hemorrhagic, con- solidated lung showed against a paler background of nonelastic emphysematous pulmonary tissue.   

Only one case of lobar pneumonia occurred. This was caused by a Type I pneumococcus. It appeared during the first days of the epidemic, and its occurrence was undoubtedly fortuitous. Four cases showed the changes of an interstitial bronchopneumonia. Here the white, thickened, pus-filled bronchioles surrounded by hemorrhagic, edematous or indurated areas of consolidation presented a very distinctive picture, entirely different from the great majority of the pneumonia processes seen.   

Microscopically, various pulmonary changes were found, but nothing was encountered beyond what one is accustomed to expect in bronchopneumonia. The accumulations of inflammatory cells either were patchy or so massed as to present, in a limited number of sections, an appearance indistinguishable from lobar pneumonia. The exudate was either very dense or thin; in either case edema and congestion were marked, thus greatly compromising the remaining lung tissue. As a rule, polymorphonuclear leucocytes greatly predominated in the exudate. In other sections, however, the presence of many mononuclear


wandering cells and epithelial cells were noted. Often great quantities of erythrocytes were massed together in the air cells, sometimes presenting a very striking picture, especially when contiguous to the other air cells stuffed with the inflammatory exudate. Fibrin was not a prominent part of the exudate except in one case. In a few cases, wide stretches of exudate appeared, from which all remains of alveolar walls had disappeared. In several such cases, focal areas of disintegration were in process, evidenced by poorly stained, fused masses of cellular debris in which many remains of destroyed nuclei appeared. The terminal bronchioles usually contained the same type of exudate described in the alveoli. The mucous membrane was often separated off, and tangled fragments appeared free in the exudate. Smears of the lungs usually showed enormous numbers of Gram-positive diplococci. It was rare to find organisms arranged in chains. Microscopic study of the remaining organs did not lead to any noteworthy disclosures. Eighty of the 123 cases of pneumonia examined post-mortem showed an associated pleurisy; 34 were described as serofibrinous, 25 as serofibrinopurulent, and 21 as fibrinous.   

Strange as it may seem, no difference was to be made out in the nature of the process caused by the streptococcus, pneumococcus, or influenza bacillus. It was not difficult to distinguish Type III pneumococcus by the sticky exudate and the greater tendency toward confluence, but no matter what the infecting organism, each appeared to produce the same type of pulmonary lesion with equal facility.
There were no particularly constant extrapulmonary complications. Nine cases showed hemolytic jaundice, the majority of these appearing late in the epidemic. Four cases showed a massive interstitial emphysema, probably beginning in the multiple rupture of the emphysematous alveoli at the pulmonary hila. Four cases of pneumococcus meningitis developed in the course of the disease, only two of which came to necropsy. There were relatively few splenic tumors; a firm, congested organ not a great deal larger than normal was the rule. Acute changes in other parenchymatous organs were inconspicuous. Marked acute nephritis occurred only three times. Rupture of the rectus muscles was found in four instances. The bone marrow of the femur in all cases studied appeared, to gross examination, pale and unreactive.   

There was a noticeable and instructive change in the course and manifestations of the disease in the last days of the epidemic. The leucopenia disappeared in large measure; 20 of 35 leucocyte counts averaged 20,000 cells per cubic millimeter. A far higher percentage of blood cultures showed positive results. Of the total of 11 positive findings, 7 were obtained in the last 80 cultures made. At necropsy the following synchronous changes were revealed: Older processess in the chest, more fibrin and pus in the pleural cavities, and collapsed lungs showing various stages of resolution and extension of the bronchopneumonia.   

The important complications in 123 cases of bronchopneumonia as disclosed at necropsy were 74 cases of infection of the pleura (serofibrinous, 34; serofibrinopurulent, 20; fibrinous, 20); lung abscess, 5; interstitial emphysema, 4; pneumococcus meningitis, 2; acute serofibrinous pericarditis, 3; rupture of the rectus muscle, 3.


The average leucocyte count in the cases of streptococcus pneumonia of the winter of 1917-18 was 17,000, and positive blood cultures were obtained in 20 percent of the cases. The duration of the pneumonia cases in the 1918 epidemic to the time of death was noticeably shorter than among cases in the previous winter. There is possible significance in the fact that the few cases of pneumonia of the 1918 epidemic that presented a picture at necropsy most like those of the winter of 1917-18 were those which had a period of longest illness, in this respect also corresponding to the cases of pneumonia which occurred during the winter of 1917-18.   

From this study of the bacteriology of the deaths of January and the spring of 1918 it appears that the streptococcus played a very important role at this time. The high death rate and case fatality rate of September and October, 1918, were apparently accompanied by a preponderance of streptococcus in the cultures, with an usually large proportion of the nonhemolytic variety. Type II pneumococcus also appears to have had a rather high incidence. The pathology indicates that the fulminant cases of short duration were affected with a hemorrhagic more or less generalized lesion with areas of spreading bronchopneumonia. Only four cases of the typical interstitial reaction are reported and occasionally cases of the lobular variety were seen. In the four cases available for study from the influenza epidemic, all showed hemorrhagic lung, the types of lesion varying in different portions of these organs. Confluent bronchopneumonia spreading out from the bronchial tree is found in all, although lobular consolidations, interlobular lymphangitis and focal areas of hemorrhages and necrosis can be found in certain portions of the lungs in each one of the cases. In two cases, in which the tissues are well fixed, Gram-negative minute bacteria are found along the air passages, but not in the tissues. These are always accompanied by Gram-positive cocci of varying morphology.   

In this camp the streptococcus was predominant throughout the entire period of the war, with an unusually large number of the nonhemolytic variety, which here, as in other camps, were found in the cases of confluent bronchopneumonia as well as occasionally in cases of the interstitial variety.   

The case fatality rate for all respiratory diseases was 5.38 percent in September, 1918, rising to 8.34 percent in October; the infiuence of the streptococcus increased as the epidemic progressed and it was responsible for the increased case fatality rate during the latter part of the epidemic.


Chart V indicates that, whereas the incidence of acute respiratory diseases at this camp, shortly after mobilization, was high, the mortality was relatively low, the case fatality rate for the month of December, 1917, being 0.21 percent, and for January, 1918, 0.8 percent. There was a sharp rise in the case fatality rate in March, the rate being 1.07 percent, and increasing to 1.97 percent in the month of April. This increase in deaths is ascribed to primary

j Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1917, 1918, and 1919.

43 CHART V.- The incidence and fatality of the acute-respiratory diseases at Camp Devens


pneumonia. The character of the cases, as shown in the histories and in the protocols of necropsies, makes it quite probable that an influenzal wave occurred at this time.   

In this camp, as in several others, a rise in the total case fatality rate of all respiratory diseases preceded the wave of influenza. The case fatality rate was 0.61 percent in June, 1918, rising to 1.48 percent in July, and to 3.83 percent in August, while the definite wave of the pandemic is not recorded as having started until September. In September the case fatality rate reached 5.48 percent, with 13,733 cases of acute respiratory disease recorded. The rate dropped sharply in October to 1.04 percent. Most of the fatal cases during the influenza epidemic were admitted for pneumonia, and the deaths are recorded in the month of September as 281 from influenza among the white personnel, while in the same personnel 48 were ascribed to primary bronchopneumonia and 340 to primary lobar pneumonia. With very few exceptions all of these cases were secondary to attacks of clinical influenza.   

During the period from September 27, 1917, to May 31, 1918, there were 485 admissions to the base hospital with a primary diagnosis of pneumonia.k The average strength of command for this period was 29,613. The 485 cases is an annual morbidity rate per 1,000 of 14.3; 64 died, or a case fatality rate of 13; there were 77 cases of empyema, with 34 deaths, or an empyema case fatality rate of 44. The frequency of empyema in pneumonia was 16 percent.   

Negroes began to arrive at this camp in number in April and were gradually increased until August when approximately 4,000 men of this race were at the camp. Those affected during the period under consideration showed 39 times as much pneumonia as did whites, while the mortality was 16 percent as against 12 percent of the whites. Pneumonia was lobar in 94 percent of the Negroes and 74 percent of the whites. Lobar and bronchopneumonia in the Negroes and bronchopneumonia in the whites showed a mortality of 15 to 18 percent, while lobar pneumonia in the whites showed only 9 percent. Empyema was one-third as frequent in the Negroes as in the whites, the Negroes apparently dying too early in the disease to develop empyema. Their mortality with empyema, however, was 89 percent as against 38 percent among the whites.   

The incidence of pneumonia decreased in proportion to the length of service, but the mortality was lower in the unseasoned men. Lobar pneumonia was four times as frequent as bronchopneumonia during this period. One must consider, however, that in a considerable proportion of these lobar pneumonia cases this diagnosis was made during the increase of the spring and that a relatively small proportion of them were autopsied. It is stated that the acute minor upper respiratory infections were disregarded because unsatisfactory histories were obtained. The diagnosis of pneumonia was made frequently by the X ray in the absence of definite physical signs. Enlarged hearts in pneumonia were also demonstrated satisfactorily by this method.   

In the typing of the organisms increased proficiency in the use of technical methods resulted in a marked decrease in the undetermined types, and suggests


k The following statements of fact are based on: Pneumonia and Empyema, by Horace Gray. Boston Medical and Surgical Journal, 1919, clxxx, 265, 305, 330, 351, 388, 422, 448, and 475.


that the bacteriological determinations for the whole series had only a relative value. In the first 100 cases of pneumonia 43 percent were reported as unknown types; in the second 100, 56 percent;in the third 100, 66 percent; the fourth 100, 38 percent; and in the last 85 cases, only 14 percent. Unsatisfactory specimens were responsible for a considerable number of the "undetermined " of the earlier findings.   

Pathologically speaking, the three main kinds of pneumonia seen in the hospital at Camp Devens were lobar, broncho, and empyema without prior pneumonic consolidation. Lobar pneumonia was nearly four times as frequent as bronchopneumonia. Measles-pneumonia was by no means always bronchopneumonia, but in 24 per cent of the cases it was lobar; of these 10 cases clinically lobar, three were proven at autopsy to be lobar, not confluent-lobular. Of all the cases of lobar pneumonia, 16 percent got empyema, as did practically the same, 15 percent, of the cases of bronchopneumonia.
The death rate for lobar-pneumonia-empyema, however, was less than half that for bronchopneumonia-empyema.   

Etiologically speaking, the three main kinds of pneumonia in this camp were primary, postmeasles, and postether. The primary made up 88 percent of the 845 cases, while only 8 percent were consequent on measles. Empyema developed in measles-pneumonia more than twice as often as in primary pneumonia.   

From the viewpoint of mortality, the highest among the pneumonias was postmeasles, 32 percent; among the primary, only 12 percent died; among the postether pneumonias 1 died, or 6 percent. Among the empyemas, the measles-pneumonia-empyemas, furthermore, had more than twice the death rate of the primary pneumonia-empyemas.
While it is clear that pneumonia here was essentially a primary disease, there were included under that heading some cases which by more refined interpretations might be called secondary, rather than associated. Among these acute minor upper respiratory infections (rhinitis, pharyngitis, tonsillitis, laryngitis, even sinusitis) were dirsegarded, because: (1) Accurate data as to their presence had not been consistently recorded; nor, indeed, is it likely that satisfactory histories could be obtained here. (2) Further, even where present, these complaints cannot be considered primary causes of pneumonia, for they have been extraordinarily frequent in healthy men. The tuberculosis examining boards found high percentages of the examinees with these symptoms but with scant signs.   

The etiology of the 77 empyemas was primary pneumonia in 88 percent as against measles-pneumonia in 8 percent. When it is recalled that primary pneumonia, in this series, was 10 times as common as measles-pneumonia, it is realized that one must consider rather the converse of the first sentence of this paragraph, namely, the fraction of the primary pneumonias to develop empyema (15 percent), compared with the fraction of the measles-pneumonias (34 percent) to do so. From this it is seen that the prognosis for measles-pneumonia is unfavorable.   

The cases of primary pneumonia (427) made up 88 per cent of the series (485 cases). Their mortality was strikingly low, only 12 percent. They de-


veloped empyema only half as frequently as did the cases of postmeasles-pneumonia, and when they did get empyema, the mortality was only half that for measles-pneumonia-empyema. Of the first 100 cases of pneumonia, 24 were due to measles, presumably because of much greater frequency of measles during that period. The largest figure in any of the four succeeding periods was 9 percent.
Measles cases developed pneumonia twice as often during the second period as during any other; in the second period, 20 percent of the measles cases got pneumonia, while the maximum in any other period was 12 percent. However, during the first period there were nine and one-half times as many measles cases, this greatly reducing the element of chance. This applies to the third period in which measles did not occur as an etiological factor in a single instance.   

Measles-pneumonia cases developed empyema twice as often during the first period as during any other.
In the 485 cases, the distribution of the pneumococcus was as follows, the organisms being isolated from the sputum: Type I, 34 cases with 4 deaths, and 2 empyemas with 1 death; Type II, 54 cases with 7 deaths, among which there were 2 empyemas with 1 death; Type III, 38 cases with 5 deaths, 2 empyemas with 1 death; Group IV, 80 cases with 5 deaths, in which group there were 5 empyemas with 2 deaths; type undetermined, 7 cases with 5 deaths. This gives a pneumonia mortality for this group of 12 percent, and there was empyema incidence of only 10 percent with a mortality of 52 percent. Cultures in which the streptococcus was found with the pneumococcus occurred in 14 cases with 5 deaths, a mortality of 36 percent, 13 of the cases having empyema with a mortality of 4, or 31 percent. Streptococcus without pneumococcus occurred in 43 cases with 21 deaths, a mortality of 49 percent; 41 empyemas with 19 deaths, a mortality of 46 percent.   
The organism in pneumonia was most often unknown (in 44 percent), then pneumococcus alone, i. e., without streptococcus (in 44 percent), then streptococcus alone (in 9 percent), lastly, a mixture of streptococcus and pneumococcus (in 3 percent).   

The mortality of pneumonia was greatest when the organism was streptococcus alone, 49 percent; then from mixed streptococcus and pneumococcus, 36 percent; then from pneumococcus alone, 12 percent; and least when the organism was unknown, only 6 percent. The most frequent organism in empyema was streptococcus alone, occurring in 53 percent of the 77 cases. Next came pneumococcus alone, 27 percent; mixed streptococcus and pneumococcus in 17 percent, and finally unknown organisms in the remaining 3 percent. The mortality of empyemas differed from that of the pneumonias in general by being highest from the pneumococcus alone, 52 percent. Next in fatality came streptococcus, 46 percent; and mixed, 31 percent. This is in striking contrast to the figures obtained by an earlier analysis of the empyemas during the first six months, 53 out of 241 pneumonias. Among these empyemas the mortality was exactly reversed, i. e., highest with mixed pneumococci and streptococci, 66 percent; then with streptococci alone, 49 percent died, while with pneumococci alone, only 38 percent. The explanation is that the first series,


the white men, did not die so readily unless the streptococcus was superimposed upon the pneumococcus, while the Negroes later died with the pneumococcus alone.  
The improvement in the number of cases typed is striking, toward the latter part of the period reviewed. The percentage of cases whose bacteriology was "undetermined" in the first 100 cases of pneumonia was 43; in the second 100, 56; third, 66; fourth, 38; and in the last 85 cases of the series, only 14. This satisfactory diminution in the number of untyped pneumonias was largely due to the growth of a general appreciation of: (1) The importance of coughed lung sputa, not hawk-up throat mucus; (2) the surprisingly large number of pneumonias from which such satisfactory specimens were obtainable only after insistence on an early morning coughing spell; (3) the kind of sputum likely to prove serviceable for typing, not easy to describe, but thin rather than thick.   

The bacteriology in the 148 cases of pneumonia in Negroes compared with that of the 337 in whites showed that the organism was streptococcus and pneumococcus in 1.4 percent of the Negroes and in 3.6 percent of the whites; streptococcus alone in 1.4 percent of the Negroes and in 12.2 percent of the whites; pneumococcus alone in 66.1 percent of the Negroes and 34.1 percent of the whites, the balance being undetermined. The typing of the pneumococcus alone in the 148 Negroes showed Type I, 16 cases, with 3 deaths; Type II, 33 cases, with 6 deaths, Type III, 20 cases, with 2 deaths; Group IV, 26 cases, with 1 death, the balance being undertermined. In the 337 whites, pneu- mococcus alone, was found: Type I, 18 cases, with 1 death; Type II, 21 cases, with 1 death; Type III, 18 cases, with 3 deaths; Group IV, 54 cases, with 4 deaths. In 97 whites in whom pneumonia occurred at the same time as in the 148 Negroes, pneumococcus alone was found; Type I, 9 cases, with 1 death; Type II, 10 cases, with no deaths; Type III, 11 cases with 2 deaths, Group IV, 11 cases, with 2 deaths, the balance being undetermined.   

It was felt that the bacteriological determinations could not be compared satisfactorily with the types of pneumonia present in the cases. This is undoubtedly true, as the only satisfactory cases from the standpoint of pathology would be those examined post mortem, and in some of these cases the bacteriology was either lacking or inadequate for the purpose. Streptococcus was definitely more prevalent among the whites.   

The complications noted were renal in 55 percent of the 485 cases, empyema in 27 percent, pericarditis in 3 percent, otitis media in 2 percent, while pulmonary abscess, jaundice, and relapse were present in 1 percent. Thirty-three percent of the cases of pneumonia showed albumin; 14 percent more had casts without blood, and another 8 percent had red blood corpuscles. Thus less than half the cases escaped kidney involvement as demonstrated by examination of the urine.   

Purulent peritonitis was found in one case in which there was empyema in the left pleura. Pleural effusions with negative cultures not diagnosed as tuberculosis were present in appreciable numbers. One patient developed empyema on the side opposite the pneumonic consolidation, with pneumococcus in the culture, while two cases of empyema developed without preceding consolidation having been demonstrated. Though it may be considered probable


that pneumonic consolidation of some kind existed in these cases, it was entirely within the realm of possibility that infection spread from the involved mediastinal lymphatics directly to the pleura without causing pneumonic consolidations in the lung. In one case an effusion occurred a week before consolidation was demonstrated.   

The failure to demonstrate pneumonia at autopsy, even though it had been clinically present during the course of the illness, though infrequent, did occur and was reported from many camps and towns of the United States as well as abroad. Five cases of this variety were reported at Camp Devens, but the deaths occurred in such stage of the disease that the presence of pneumonia at some time during the course of the illness can not be ruled out. Three of the cases had relatively marked empyemic involvement.   

It was frequently impossible to diagnose definite pneumonic consolidation on the basis of physical signs. A large proportion of the cases were diagnosed lobar pneumonia but all types of bronchopneumonia are described, while the autopsy protocols indicate that in a high percentage of cases in which diagnosis of lobar pneumonia was made, there was really mixed pneumonia, with varying reactions in different lobes of the lung, though massive consolidations were frequent. With the high incidence of the fixed types of pneumocci it is to be expected that the lobar consolidations would be found frequently at least in some of the lobes and this was evidently true, yet it was relativley rare that clear-cut lobar pneumonia, without other consolidation in the lung, was described. The classification of such cases is based on the predominance of one type of lesion and in that sense only does true lobar pneumonia predominate, as indicated by the necropsy protocols.   
The bacteriology of the influenza epidemic at this camp was carefully studied.l  
In addition to sputum typing, smears were axamined and careful cultures were made using, for a while, blood on agar plates, and later, 10 percent defibrinated blood agar plates. Symbiotic action of staphylococcus and B. subtilis was used in some instances to enhance the growth of the influenza bacillus.   

The presence of B. influenzae in the smear alone was considered inconclusive and its location in pure culture was considered necessary. From the sputum, B. influenzae was cultivated 104 times, 12 times in pure culture and 92 times in association with other organisms. In the 92 cases there were 5 Type I; 2 Type II; 1 Type II atypical; 5 Type III; 52 Group IV; not typed, 25. Cultures of the pleural fluid showed pneumococcus Type I, 11, Type II (subgroup), 5; Type III, 2; Group IV, 7; Streptococcus hemolyticus, 3; B. influenzae (pure) 2; B. influenzae and Streptococcus hemolyticus, 2; B. inftuenzae and pneumococcus, 4; no growth (serous), 8. B. influenzae was not isolated from the blood cultures. The organisms found in the blood culture were pneumococcus, Type I, 10; Type II, 7; Type III, 1; Group IV, 9; Streptococcus hemolyticus, 1. The high incidence of pneumococcus Type I and II is noteworthy. Cultures at autopsy in 37 cases showed B. influenzae predominating in 14; B.


l The following statements of fact are based on: A Bacteriologic Study of the Influenza Epidemic at Camp Devens, Mass., by Lesley H. Spooner, Joseph M. Scott, and Elmer El. Heath, jr. Journal ofthe American Medical Association Chicago, 1919, lxxii, No. 3,155-159.


influenzae and pneumococcus, 4; B. influenzae and Streptococcus hemolyticus, 3; B. influenzae and Staphylococcus aureus, 1; B. influenzae, Streptococcus viridans and pneumococcus, 1; Streptococcus hemolyticus, 1; pneumococcus, 6; negative, 7.   

B. influenzae
occurred alone or with other organisms in 23, or 62 percent, of 37 cases. In addition, B. influenzae was recovered from the frontal sinus in 4 cases; from the sphenoidal sinus in 4, and from the middle ears in 2.   

Although the B. influenzae was present in frequency enough to warrant the statement that it was the most important etiologic factor of this epidemic, vet considerable importance must be placed on the secondary invaders, the pneumococcus, and in this hospital rarely the Streptococcus hemolyticus, which were found in such a large percentage of cases examined, both during life and post mortem.   

Agglutination reactions of cultures of B. influenzae were made with patients' serum at various periods after the onset of the disease.   

From these studies it seems reasonable to suppose that the prime factor in this epidemic was B. influenzae since it was found in such a large proportion of specimens of sputum when the latter was derived from the lower air passages and was properly examined; since the organism was recovered from lungs post mortem in 62 percent of the cases carefully studied, and in pure culture from at least one lobe in 50 percent of the same series; and since the blood of patients convalescent from the disease showed a rising agglutinating power not only to their own organism, but also to heterologous cultures.   

The direct smear method was of value in indicating what might be expected in culture. The most striking fact in regard to such preparations lies in the great frequency of intracellular B. influenzae. The absolute necessity of immediate necropsies, if the influenza bacillus is to be cultured before overgrown by other organisms, was convincingly demonstrated.   

Interesting studies were made of the pathology of the fatal cases of influenza in the pandemic.m Two types of lungs were noted as strikingly characteristic findings in this disease. The first was encountered in cases in which death occurred within a few days after the onset of pulmonary signs. These cases yielded lungs which were partially collapsed, dark red, lax, but meaty in consistency. The pleural surfaces were often partly covered with a dusky red mottling, due to small extravasations of blood beneath the serous coat. In some cases there was a thin layer of dusky red fibrinous exudate upon the pleural surfaces, particularly over the posterior borders. On section these lungs were dark red and wet. They were dripping wet, and the fluid from some portions was a blood-tinged serous liquid and from others dark red and bloody. On close inspection the cut surfaces usually were found to be thickly sprinkled with air vesicles of considerable size. The lung tissue as a whole, after the liquid had drained from it, was brownish-red in color, and somewhat translucent and friable. The mucosa of the bronchi was usually very dark red in color, and the bronchial lymph nodes were enlarged and deep red in color.   

The other type of lung, which was found in patients that had lived for 10 days or more after the onset of the disease, while showing traces of the type of

 m The following statements of fact are based on: Comments on the Pathology and Bacteriology of Fatal Influenza Cases as Observed at Camp Devens, Mass., by S. Burt Wolbach. Johns Hopkins Hospital Bulletin, Baltimore, 1919, xxx, 104-109.


lesion just described, was characterized by a very extensive bronchitis, with bronchopneumonia, discrete or confluent, and peribronchitis. Such lungs were more voluminous than the preceding, but they did not fill the chest cavity post mortem. They were nodular, and the pleural surfaces occasionally showed a striking tracery, due to the injection of the subpleural lymphatics. Portions of the surfaces of the lungs, in some instances, were covered with a thin layer of fibrinous exudate. On section, the most prominent feature was the extensive injection of the bronchi, particularly the smaller ones, with a fibrino-purulent exudate. The injection of the bronchi at times was so extensive and uniform as to produce geometrical patterns, which were very striking when the condition was accompanied, as it usually was, by a marked infiltration of the interlobular septa. A casual inspection sufficed to show that the smaller bronchi were dis- tended, usually markedly dilated, and in cases of two weeks' duration spherical and cylindrical bronchiectases were very common. The condition, in fact, was one of panbronchitis; peribronchitis with extensive infiltration of the interlobular septa; and organization in alveoli and bronchioles.   

In the first type of cases, in which emphasis was laid on the gross appearance of the lungs, B. influenzae was the only organism which could be cultivated, and these distinctive conditions were unhesitatingly associated with that organism. In the lungs showing other types of solidification, other organisms were responsible for the exudation characterizing the pneumonias. The hemolytic streptococcus, the staphylococcus, and the pneumococcus, each produced its distinctive picture, the last often that of lobar pneumonia.   

While the bacteriological evidence, based upon the assumption that B. influenzae is the cause of influenza, was very good in support of the stand that there was a distinctive lung lesion in these influenza pneumonias, the histological study afforded very definite proof. Early in this study of the Camp Devens cases, the fact was recognized that a striking type of reaction was present, a condition of acute alveolar emphysema with the deposit of a hyaline fibrinous material on the alveolar walls. The intervening alveoli were compressed and filled with exudate, which in the early cases was largely serous or bloody, containing but little fibrin. It was this acute alveolar emphysema, with the serous and hemorrhagic exudate, that gave the characteristic gross appearance to the lungs in the early stage of the disease. In order to determine how com- mon this lesion was all of the General Hospital No. 10, Boston, autopsies on influenza cases were studied, and it was found to be constant. It might be masked by a pneumococcus or streptococcus exudation or by extensive hemorrhage, but its presence could always be determined by the finding of the hyaline fibrin outlining greatly distended air spaces in the lungs. It was the one distinctive feature in the pathology of influenza pneumonias, and its constant occurrence was indicative of the entity of the initial lung infection. The interpretation of this lesion was not easy. The hyaline fibrin, because of its prominence and the juxtaposition of cellular exudate, often simulated the outlines of alveoli. As a matter of fact it outlined cavities filled with air which might or might not completely fill groups of alveoli. Although alveolar walls in contact with this fibrin might be necrotic, tissue elements played no part in its formation. A similar hyaline fibrin was found in two cases of emphy-


semia of the mediastinunm where the mediastinal areolar tissues were infected by pneunococcus, secondary to pneumococcus pericarditis. The physical characteristies of this fibrin were determined by its contact with air, and an important factor was probably the mechanical compression of strands of fibrin by air. What is the source of the exudation in the alveoli in these early pneumonias? The exudation might be present in alveoli with intact walls, or walls showing very slight reaction, mainly evidenced by activity of the respiratory epithelium. In all cases severe lesions were found in the finest bronchioles and in the alveolar ducts. The latter showed an exudation composed mainly of polymorphonuclear leucocytes and small quantities of fibrin. The walls were filled with leucocytes, and often were necrotic in places. The intralobular bronchioles showed severe lesions of the mucosa, and often it was possible to demonstrate the source of hemorrhages from capillaries. The obvious expla- nation, and indeed the only possible one from the material studied, is that the major injury was to the bronchial system, and mainly in the finest bronchioles and alveolar ducts. To secure the degree of emphysema present it was necessary to assume a valve action of the exudate in the bronchi. The character of the hyaline fibrin deposit around air vesicles upon the alveolar walls suggested a pouring of exudation into the alveoli from the bronchioles and alveolar ducts, at a time when air was able to pass. Thus the patient was virtually blowing bubbles in his own lungs, into a medium of exudation relatively poor in fibrin.   

The mechanism of interstitial emphysema formation was easily seen, where the greatly distended alveoli were in contact with the pleura of interlobular septa. In these locations it was possible to demonstrate rupture of the alveolar walls and the direct continuity of fibrinous strands, partially filling clefts dissected by the air from alveoli to plueral or interlobular connective tissue. A series of gross sections and microscopic sections from lungs with interstitial emphysema showed that the air found the easiest route of exit from the lung in the connective tissue surrounding blood vessels. It dissected along blood vessels to the hylus of the lung and from there along the great vessels and bronchi into the mnediastinum, over the pericardium into the anterior mediastinum, and upwards along the trachea into the tissues of the neck, whence it escaped into the subcutaneous tissues. This subcutaneous emphysema might appear very early. The earliest case was seven days from the first symptom, which meant, of course, a shorter duration of the lung involvement. The majority of the cases were noted on or after the tenth day from the initial symptoms of the disease.
It must le borne in mind, in considering the pathology of these lungs, that the lesions were not uniformly distributed, and therefore very extensive injury in portions of one or several lobes was compatible with life for a considerable period of time. The bronchial lesions apparently progressed, and might extend throughout the whole of one or both lungs, producing the anatomical picture of the more chronic cases, that of a panbronchitis with bronchiectases and peribronchitis. During this period of extension in bronchi, a number of things might happen to the portions of the lungs first involved. They might become


secondarily infected with pneumococcus or streptococcus, or the Gram-negative diplococcus, called by English workers, "Diplococcus mucosus." In rare instances staphylococcus and Friedlander's bacillus were encountered. The fate of the tissue depended on the nature of the infecting organism; as, for example, fibrinous exudation with the pneumoccus and abscess formation with the staphylococcus. In a number of instances these portions of the lungs, severely damaged at the onset, did not become secondarily infected; at least, these lungs showed only the influenza bacillus at the autopsy, and had undergone extensive organization resulting in cicatrices of large sizes. If one take a series of lungs which have shown only the influenza bacillus in cultures and in sections, one may still have all the stages described, exclusive of those with secondary infection, and accordingly it must be concluded that the reaction to the influenza bacillus is less intense in the later stages of lung involvement than in the earlier. This was shown best in comparing two lungs front the same case, where in one lung, usually the right, there was found the severe damage of the early lesion with bronchiectasis and peribronchitis, and in the other luger a much less intense bronchial reaction, with much less marked peribronchitis, or none at all. The involvement of the pleura in lungs infected solely with influ- enza bacillus was very slight. There were hemorrhages into the pleura and perhaps a thin layer of fibrin upon the surface. The amount of fluid in the pleural cavities was always small, though blood-tinged. Empyema was found in cases secondarily infected with the streptococcus or pneunmococcus. The involvement of the pleura sometimes results from the extension of the inflammatory process along the interlobular septa and lymphatics, or, as was more commonly the case, from bronchiectatic cavities situated close to the pleural surface.   
Gangrene of the lung was noted in one of the Camp Devens series in a case showing very extensive bronchiectases, with bronchiectatic abscesses. Extensive necrosis of the lung was observed in a number of cases in this same series and at General Hospital No.10 necrosis due to organisms other than the influenza bacillus.   

Organization in the pure B. influenzae cases was a common end result. The organization of the exudate began early, certainly before the tenth day of the disease, and a prominent factor in bringing about this result was believed to be the plugging of the bronchi with exudation. In patients who had survived three weeks or more there were very complicated gross appearances, due to extensive cicatrization of large portions of the lung. The contraction of the interlohular septa, due to the avascular organization of exudate, ecaused marked distortion of the lobules of the lung, and peculiar lines of retraction on the pleulral surfaces.   

While it is not the purpose here to include the whole pathology of influenza, a few interesting features in other organs are worthy of emphasis. Eight of the Camp Devens series showed waxy degeneration of the rectus museles, and subsequent experience at General Hospital No. 10 indicates that it was probably overlooked in some of the earlier post-mortem examinations made at Camp Devens. A number of these cases showed rupture and extensive hemorrhage into the rectus muscle. This lesion was noted in other muscles; for instance,


the transversalis, the internal and external oblique muscles, the latissimnls dorsi, the pectoralis major and the intercostal muscles. The testes occasionally showed minute petechiae, but on the whole no striking gross change was observed. Microscopically very striking changes were encountered in nearly every case, namely, the cessation of activity in the seminiferous tubules; actual degenerative changes frequently were noted, and in late cases beginning fibrous tissue replacement of the degenerated tubules. This lesion of the testis seemed to be wholly a toxic one, as there was very little cellular reaction. It was difficult to understand why such severe toxic lesions of the niuscle and testes should occur, in the absence of effects attributable to toxins in other organs. For instance, the reaction of the spleen was very slight, the heart muscle rarely showed any gross or microscopic lesion, and in general seemed to escape entirely the toxic effect of the disease. Lesions of the adrenal, when extensive, such as hemorrhage, could be attributed to secondary infection, usually the hemnolytic streptococcus. Minor acute lesions constantly were found in the cortex in influenza cases, but these lesions were similar to those found in many infectious diseases-the disappearance of lipoid content, and focal necrosis with mononuclear phagocytic cell reaction. The head was opened in 20 of these cases. Infection of the middle ears was found in 13, infection of the sphenoidal sinus in 20, frontal sinus in 7, and of the ethmoidal cells in 8 cases. Three cases showed punctuate hemorrhages in the cerebral cortex.   

From the foregoing studies it is seen that the early pneumonias of this camp following measles, primary, and during the apparent influenza rise in the spring, were characterized by a relatively large number of cases which bacteriological examination indicated were due to infection with the fixed types of pneumocoeci, with a high mortality from both pneumococcus and streptococcus infections. The type of reaction in the lungs was either lobar or confluent bronchopneumonia, with a relatively small number of the suppurative variety, in which Streptococcus hemolyticus was most frequently found. The relatively few eases in the influenza epidemic of September, 1918, suggest B. influenzae as the causative organism of the primary infection. At least, it was the earliest and most frequent of all bacteria discovered.   

Necropsy protocols for 229 cases of pneumonia, with 19 specimens, were received at the Army Medical Museum from this camp. As previously stated, the type of the pneumonic consolidation showed a large percentage of fibrinous consolidation, usually, however, accompanied by other types of consolidation in other portions of the lung. Lobar types were apparently much less frequent during the influenza epidemic. The picture given above of the lesions accompanying infection by the influenza bacillus is recorded in the protocols as having been present in most if not all of the cases during the epidemic period though frequently masked by other types of lesion. The bacteriological picture as shown by stains of the fixed tissue is rarely satisfactory unless autopsy followed death within a relatively few hours. Throughout the duration of this camp Streptococcus hemolyticus did not play an important part. The fixed types of pneumococci appear to have been largely responsible for the secondary pneumonias.


CAMP DIX, N. J., n

Measles and common respiratory diseases assumed epidemic proportions shortly after mobilization began in September, 1917. Relatively few deaths occurred and the case fatality rate was low, as is indicated in Chart VI. Influenza was diagnosed in considerable numbers throughout the fall of 1917. The sharp rise in respiratory diseases, with peak in March, 1918, was due to an epidemic of influenza, over 1,100 cases being so diagnosed in that month. The case fatality rate, however, was low. The deaths were ascribed to primary pneumonia for the most part, the preliminary respiratory infection not being considered as a diagnosis, as is true of most of the other camps. In a general way the pneumonic lesions in the spring of 1918, appear to have been, in large part, bronchopneumonia, and the streptococcus does not appear to have been very frequent, though it was observed, particularly in the empyema cases. The low case fatality rate suggests that the organisms were not particularly virulent during this period.   

Thorough clinical and laboratory studies were made during the influenza epidemic at Camp Dix, which began September 15, and ended October 6, 1918.o During the 22 days of the epidemic, 6,500 patients were cared for. Approximately 6,000 of these men had influenza; 800 deaths were due to the epidemic.   

The usual history was one of gradual onset with prodomes for four or five days prior to admission, consisting of headache, malaise, backache, myalgia, fever and chills or chilly sensations, and marked prostration. There was a history of anorexia, and of aggravation of symptoms, after drill or exercise. Occasionally the onset was sudden, sharp, and severe. A low temperature in a severe case was an unfavorable sign. The blood count usually showed leucopenia. The urine contained albumin and casts. The patients looked very ill, but often did not feel so. Few complained of sore throat or the early sticking pains in the chest, aggravated by coughing, so commonly observed in beginning pneumonia.   

One-third of the cases presented frank signs of pneumonia. Every patient who had fever, prostration, rapid pulse, increased respirations, cough, and bloody sputum was at once put down as a pneumonia suspect, which diagnosis could usually be confirmed later by physical signs and roentgen examination.   

Frequently no physical signs of pneumonia were apparent on the initial examination. A day or two later, bronchopneumonia was discovered, the lobules subsequently tending to become confluent. This condition sometimes gave the physical signs of a lobar pneumonia, but the real condition was repeatedly demonstrated at necropsy. The roentgen ray was an invaluable aid in the diagnosis of such cases.  
The outstanding feature of the disease was the extreme toxemia noted in the serious cases. Almost from the first inspection the outcome of each case could be predicted. There was a sharp line of demarcation between the serious and mild cases. In the former many patients developed cyanosis early in the disease and died promptly, after periods varying from a few hours to three days after admission.
n Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1917, 1918, and l9l9 o The following statements of fact are based, in the main, on: The Influenza Epidemic at Camp Dix, N.J., by Martin J. Synnott and Elbert Clark. Journal of the American Medical Association, Chicago, 1918, lxxi, No. 22, 1816-21.

55   CHART VI.- The incidence and fatality of the acute respiratory diseases at Camp Dix


The cyanosis appeared early and was progressive. It was a fairly constant and characteristic symptom in the severe cases. This intense cyanosis was a striking phenomenon. The lips, ears, nose, cheeks, tongue, conjunctivae, fingers, and sometimes the entire body, partook of a dusky, leaden hue. Frequently this cyanosis was apparent before there were any demonstrable physical signs of pneumonia. The cause could not be ascertained, a few spectroscopic readings failing to show absorption bands of methemoglobin. As necropsy revealed marked so-called compensatory emphysema in these cases, and as cyanosis is a common accompaniment of emphysema, possibly this was in some way accountable for the condition. Otherwise a purely mechanical conception of the mode Gf production of the cyanosis seems untenable. Lividity, occurring early in the disease, proved to be an ill omen.   

Very often this cyanosis came on suddenly in patients who had been doing well. It was not due to cardiac dilatation. The pulse was often slow, full and regular in such cases, and remained so, with a rate under 100, almost until death.  
The disparity between temperature and pulse was striking. Patients with temperature between 104° and 106° F., not infrequently had a pulse rate below 80. This was regarded as a strong point in differential diagnosis between these influenzal pneumonias and those of pure pneumococcic origin.   

There was noted in many instances a distinct tendency to relapse. In several mild cases, after a day or two of nearly normal temperature, exacerbations occurred and the infection changed to the severe, toxic type.   

Herpes labialis was relatively infrequent, but became more common as the disease advanced. It was seemingly of little help in prognosis, as many with herpes died, contrary to the teachings of the older clinicians.   

The sputum showed varying characteristics. It was mucoid, mucopurulent, blood streaked, frothy and bloody, or bloody as in infarct. In one it was thin, brown, without froth, and homogeneous (prune juice variety). At times it was rusty but lacking tenaciousness and of the consistency of typically pneumonic sputum; it was odorless. In one case the patient coughed up a large quantity of homogeneous, thick, purulent, greenish sputum at one time. Necropsy in this instance revealed multiple bronchopleural fistulas leading to an encapsulated empyema on the right side.   

Flushing of the face was common. An early generalized erythema resembling that of scarlatina, most noticeable on the chest and back, was often observed. Miliaria (subamina) was common later in the disease, especially in lethal cases in which sweating was frequent and excessive.   

The throat was usually infected, but not painful; the tonsils were not swollen as a rule; there was marked conjunctival congestion at the outset of the disease. The tongue was variable; usually coated, moist in mild eases, dry, cracked and occasionally bleeding in severe cases. Sordes appeared frequently on the teeth.   

Hoarseness was frequent, due either to the trauma of coughing or to superficial laryngeal ulceration or edema, as demonstrated at necropsy.   

The mental condition either was apathetic or there was an active delirium. In moribund patients, motor and psychic restlessness was remarkable. In


some cases the typical typhoid state existed; stupor, low muttering delirium, subsultus tendinum, carphology, incontinence of urine and feces.   
In one case necropsy revealed the presence of bilateral abscess in the lower rectus (Zenker's hyaline degeneration).   

There were fewer cases of enipyema in this epidemic than might be expected, owing to the extreme toxicity of the disease, with the result that death occurred early. Most of the cases examined post mortemi showed lesions in the pleura which undoubtedly would have resulted in empyemna had the patients lived longer.   

One of the most striking of the complications was hemorrhage from the mucous membranes, especially from the nose, stomach, and intestine. Frequently, pneumonia patients would have hemnoptysis like the hemorrhage of tuberculosis. Bleeding from the ears, and petechial hemorrhages in the skin also occurred. Purpura was seen rather frequently. Other complications were phlebitis, thrombosis, or embolism of peripheral arteries (with gangrene of the foot in two cases), toxic erythema, vomiting, diarrhea, conjunctivitis, convulsions, purulent peritonitis, inflammation of the accessory sinuses and of the middle ear, and pericardial effusion. A pronounced jaundice, not obstructive, as the stools were not acholic, and probably of infectious origin, was noted in many of the severe cases. Retention of urine was not uncommon.   

In several instances the infection produced a pronounced hemolytic effect, with rapidly progressive anemia. In one instance, that of a medical officer, the red corpuscle count was reduced to 1,600,000 with 50 percent hemoglobin on the fifth day of the disease.   

The pathology of the cases of influenzal pneumonia coming to necropsy at Camp Dix presented very diverse pictures. In the cases seen at necropsy the lesions were confined for the most part to the chest cavity. This was especially true of the cases which ran a rapidly fatal course. In the latter, the greater part of the visible pathologic changes consisted of very much congested and hemorrhagic, water-logged lungs.   

A notable feature of the cases in which death occurred early in the disease was the extreme water-logged appearance of the lungs-hlngs that were almost, if not completely filled with a watery, bloody and frothy fluid, with petechial and large hemorrhagic areas in the pleura. The right lung in one ease weighed nearly 4½ pounds and was devoid of air except for a small area at the apex of the upper lobe and small areas here and there along the anterior margin. The latter areas appeared to contain air under pressure. The remainder of the hlng was tightly filled with fluid, was soft at all points, arid did not contain any nodules or firmly consolidated areas. When the lung was held up, large quantities of fluid ran out. A slice cut from the lung lost about half its weight in fluid squeezed out by pressure with the hand. From this lung there was grown a nonhemolytic streptococcus and the influenza bacillus. Pronounced inflammation of the trachea and bronchi was noted in all cases in which death occurred early in the disease.   

While only one case of well-marked empyema came to necropsy several showed evidence of beginning empyema.


The heart showed nothing beyond dilatation of the right ventricle in the more acute cases. Early pericarditis was seen in cases of longer duration. The spleen as a rule showed no evident change. The liver weighed on the average approximately 2,000 gin. and was congested and often yellowish. The kidneys showed a markedly congested and yellowish cortex. The intestinal tract showed in the majority of cases, only slight congestion, and petechial hemorrhages in the gastric mucosa.   
Blood examination in more than 700 cases revealed in most instances an absence of leucocytosis in the more severe cases and during the acute stage of those running a more favorable course. The average was about 5,000 white cells per cubic millimeter during the acute stage. The lowest recorded was 1,200.   

The lymphocytes experienced a relative percentage increase. The red corpuscle count, as a rule showed no decrease and frequently a high count during the acute stage. After two or three weeks in the hospital the picture changed in most of the serious cases. The red corpuscle count decreased sometimes to 1,600,000 and a leucocytosis developed, ranging from 12,000 to 30,000 white cells, with polymorphonuclear leucocytes running as high as 95 percent. The hemoglobin varied directly with the red corpuscle count.   

The urine in practically all seriously ill patients showed a liberal amount of albumin with frequent casts.   

A large variety of organisms was encountered in cultures and smears from the lung substance, from the bronchial mucous membrane and from the sputum. Streptococci and pneumococci were most frequently found. The influenza bacillus (Pfeiffer's) was encountered in sputum, bronchi and lungs, but no particular effort was made to study it or to determine its frequency. It was encountered, however, in the majority of cases when looked for. It was recovered from the lung substance, from the bronchi, from the trachea and from the sputum, but in none of the large series (over 300) of blood cultures.   

It is worthy of note that Bacillus influenzae in no single instances was the sole invading organism. It occurred, as stated, in the sputum and in secretions taken directly from the trachea and bronchi and from the lung, but it never was in pure culture. It was always associated with one or more pathogenic organisms. It was found associated with Micrococcus catarrhalis, with the pneumococcus of various types, with Streptococcus hemolyticus or viridans, with pneumococci and streptococci, and in one or more of these combinations, plus various other undetermined organisms. Of the latter class there were observed spirilla, Gram-negative cocci and bacilli and frequently a pleomorphic Gram-positive coccus. It would thus appear that, whatever role Bacillus influenzae played in this epidemic, it did not invade the blood, and in all probability was not solely responsible for the fatal termination.   

The sputum in more than 500 cases was examined as to organisms and subjected to test for pneumococcus type by the white mouse necropsy method. The organisms mentioned above were recovered, namely, pneumococci of various types, Streptococcus viridans and hemolyticus, Micrococcus catarrhalis, Bacillus influenzae, a Gram-positive pleomorphic coccus, and Gram-negative bacilli and cocci in the associations already noted. During the early phase of


the epidemic when the pneumococcus was encountered it failed, as a rule, to give any type reaction by either the precipitin or the aggultinin test. Later, Type III and to a less extent Type II pneumococci became more and more frequent. Three hundred and fifty urines and 120 spinal fluids from seriously ill patients were subjected to the precipitin test. The results were comparable to those from the sputum examination; Types III and II became more frequent during the later phase of the epidemic. It would thus appear that neither pathologically nor bacteriologically was this a definite disease. At necropsy there were found a variety of conditions and a multiplicity of organisms in the various cases. The epidemic, however, showed phases, the later of which tended to resemble the bronchopneumonia of streptococcus and pneumococcus origin.   

From this study it is seen that, so far as bacteriology is concerned, the lesions for the cases dying early appear to have been caused by the influenza bacillus. Infection with this organism was followed by secondary infection with other organisms, most frequently pneumococci. Pneumococcus, Types II and III, apparently increased towards the la-tter part of the epidemic. These organisms, in several camps, were found more frequently as the case fatality rate increased. The case fatality rate, in September, 1918, the peak of the pandemic, was 9.65 percent, a relativly high case fatality rate, which, in the absence of any indication of a prevalence of Streptococcus hemolyticus, must be attributed, at least in part, to members of the pneumococcus group. The cases, referred to above, in which subcutaneous emphysema was noted, were made the subject of special study.P   

Twelve cases showing peculiar accumulation of gas in the fascial tissues were studied as closely as the exigencies of a crowded service would permit. Three of the cases recovered. A total of 20 was reported.   

Gas first made its appearance in the subcutaneous tissue at the base of the neck anteriorly and laterally and over the upper portion of the chest from the clavicle to about the third rib. While this was the usual location two cases showed a more extensive distribution of gas. In one case gas was found also in the subcutaneous tissue of the cheeks, eyelids, mesial aspect of the entire right arm and in small areas over the sides of the chest and abdomen and in the flanks.   

During life the patients showed a slight ballooning of the skin over the affected areas. These areas were distinctly crepitant, the gas disappearing from the immediate region of the palpating finger on pressure. A notable feature was that the skin over the ballooned area felt cold to the touch while the neighboring tissues were feverish. There was no perceptible change in color of the tissues in the affected areas.   

The patients with this accumulation of gas showed no related subjective symptoms, and no symptoms of an intercurrent infection. There was no discoloration of the affected areas. No edema, no congestion and no induration or other signs of infection were observed. It was not associated with any other

p The following statements of fact are based. in the main, on: Influenza-pneumonia Cases Showing (as in Fascial Tissues, by Elbert Clark and Martin J. Synnott. American Journal of the Medical Sciences, Philadelphia, 1919, clvii, 219-221.


especially characteristic symptoms or physical signs. The accuinulation of gas apparently did not affect the course of the disease. It appeared to be an inert gas in the tissue spaces. The affected tissues gave a crackling sound on pressure with the stethoscope. In one case there were recorded, over the precordium, rȃles very similar to sounds elicited with moderate pressure of stethoscope over emphysematous cutaneous tissue. At autopsy the mediastinal connective tissue was distended with gas in this case.   

Three of the cases came to necropsy. The pathological findings were not those of cases dead of infection with Bacillus aerogenes capsulatus, nor in fact did the lesions present the characteristics of an infection.   

In the three subjects which came to necropsy the distribution of the gas likewise varied in extent. In the most pronounced case the gas was distributed in the subcutaneous fascia as follows: Over the upper anterior portion of the chest as far laterally as the axilla, continuing upward over the front of the neck and laterally as far as the anterior border of the trapezius muscle, over the parotid region, over both cheek areas and eyelids. Similarly it could be felt in the vascular sheaths of the right axillary, brachial and radial arteries as far as the base of the thumb. Within the chest the fascia of the anterior and posterior mediastinum appeared as a reticular tissue distended with gas and bearing more or less fatty tissue. Gas was found also beneath the pleura at the hilum of the lung and in the smaller septa of the lung. In the last-named situation small gas vesicles were seen especially frequently at the edges of the small septa passing inward from these. In no case was gas observed within the muscle sheaths, and the muscles were firm and normal in appearance.   

On section the superficial gas-containing fascia of the neck appeared distinctly whitish. The blood vessels appeared to contain less blood than those elsewhere. There was no discoloration, no fluid, no infiltration or other signs of pathological change. There was no particular odor to the affected tissue and neither the liver nor any other internal organ gave any evidence of gas accumulation.   

Small bits of the gas-containing fascia were removed for inoculation on various solid and liquid media. These were incubated both aerobically and anaerobically. In no case was a growth obtained. No bacteria were found in the smears made directly from the excised and crushed gas-containing tissue.   

It is evident that air from the lung reached the pleura along the vessels and interlobular tissues and followed the adventitia to the mediastinuni. From the latter location it appears to have been distributed along the vessels to the subcutaneous tissues.   
A special board for investigating pneumonia was sent to Camp Dix during the influenza pandemic. The following is a description of one of the cases. It is typical of cases of infection with the influenza bacillus which were not primarily fatal, yet apparently were not followed by infection with other organisms.q

The pleural cavities contained no fluid, but the pleural surfaces were adherent or covered with a thin layer of fibrin. The bronchi were pale, and exuded a a thick, sticky, yellow pus. The lungs were pale, and rather distended with air,

q Pathological Anatomy of Pneumonia Associated with Influenza, by WV. a. MacCallum. Johns Hopkins Hospital Reports. Baltimore, 1921, xx, 149-249.


but throughout their substance could be felt many shotlike nodules and firm areas of rather larger size. On section the whole cut surface was pale and rather dry, and there projected everywhere firm, yellow nodules with smooth, dense, shiny surface of larger area or similar grayish yellow, firm tissue. In the center of many of the nodules could be seen the lumen of a bronchiole exuding a tiny droplet of pus.   

The bronchi, trachea, and larynx were pale, but covered with a sticky mucoputrulent exudate. This exudate, whether from the bronchioles in the lung or from the trachea, was rich in influenza bacilli. In two cases these were in pure culture, as shown by bacteriological methods and by the staining of the bacteria in sections of the lung. In several other cases the bacillus of Pfeiffer was the predominant organism, although there were other bacteria in small numbers in the lungs.   

In every case studied at Camp Dix the influenza bacillus was found either in the lungs or in the upper respiratory tract or nasal sinuses.   

Microscopically, the changes in such a lung reminded one at once of those in the interstitial bronchopneumonia caused in the former epidemic by the hemolytic streptococcus, and it is indeed an interstitial bronchopneumonia.   

The bronchiolar walls were greatly thickened by infiltration of mononuclear cells and by an enormous new formation of connective tissue. Leucocytes were strewn abundantly through these walls also, and extended into the walls of the adjacent alveoli.   

The bronchi were filled with an exudate of polymorphontclear leucocytes, many of which were loaded with influenza bacilli, while numerous bacilli were free in the fluid. The adjacent alveoli contained leucocytes, with some fibrin and epithelial cells, but no conspicuous hemorrhage. Instead, the exudate was rapidly undergoing organization, and the strands of connective tissue added to the density of the nodule. The alveolar walls were very greatly thickened, partly by the leucocytic infiltration referred to above, partly by the inwandering of mononuclear cells and the new formation of connective tissue. The walls of blood vessels, the interlobular septa and the pleura all were moderately thickened in the same way. The change differed from that produced by the streptococcus in several ways. The influenza bacilli were restricted to the bronchi, and did not appear in the alveoli, the lymphatics were quite inconspicuous, and were not found distended with thrombi laden with bacteria, and the pleura was not infected as in the streptococcal pneumonia. The absence of great pleural effusions of fluid made these cases very different from those caused by the streptococcus. Organization of the exudate was a far more striking feature in the pneumonia caused by the influenza bacillus than in the streptococcal form.   

In the collection at the Army Medical Museum the influenza pandemic at Camp Dix is represented by 48 necropsy protocols of cases dead of pneumonia, in 9 of which gross and microscopic specimens were included. The studies by Clark and Synnott, and MacCallum, and the material available at the Army Medical Museum, indicate that during the pandemic the pneumonias, as a whole, resembled those in other camps in which the streptococcus hemolytictis was not an important factor. The pneumococcic lesions of the early part of the


pandemic were massive infiltrations without definite consolidation, or with relatively lax consolidations not typically lobar in their character or distribution. Excellent cultural methods, which included taking material from many parts of the respiratory tract, resulted in the isolation of the influenza bacillus in a large proportion of the cases, together with the predominant organisms of the secondary infections. The pneumococci appeared to have had considerable virulence at this camp, particularly during the latter part of the epidemic, and it is significant that at this time there was noted an increase in Type II and Type III. Pneumococcus Types II and III which have shown, as a whole, a high mortality, gave rise to the high case fatality rates.


Camp Dodge was a National Army cantonment and received its first increment of troops between September 1 and 15, 1917. Promptly upon opening the camp, measles, together with other acute respiratory diseases, made its appearance in epidemic proportions, as is indicated by Chart VII. A peak of incidence is shown in December at which time the case fatality rate for all respiratory diseases was 0.7 percent. Two hundred and sixty-five cases were diagnosed influenza in this month, though the deaths were attributed, for the the most part, to primary pneumonia. The increase shown in the chart, with apex of incidence in March, 1918, followed an increase in personnel, but there was a marked change in the character of the pneumonias in anatomical type and in the variety of organisms which preponderated in the cultures. The case fatality rate for all respiratory diseases which had been below 1 percent rose to 2.63 per cent in the month of March and increased sharply during the latter part of this epidemic wave, the rate for all respiratory diseases for the month of April being 3.84 percent. The increased incidence of streptococcus in the cultures with resultant change in the type of pneumonia is responsible for this increased death rate. This epidemic was followed by a decrease in the incidence and fatality of all respiratory diseases, with the exception of the incidence of measles, which remained relatively high.

The case fatality rate increased from 0.64 percent in July, to 3.85 percent in August, preceding the occurrence of the pandemic of influenza. Definite influenza cases appeared in considerable numbers in September, when the case fatality rate increased to 5.40 percent and reached its peak for the war at this camp in October, when it was 6.28 percent. It dropped to 1.61 percent in November and thereafter continued to decline.   

The clinical course of the pneumonias and the pathology found therein during the fall months of 1917 and up to the epidemic of March, 1918, were not considered remarkable by the medical personnel of the camp. The majority of the cases of this period were clinically lobar pneumonia although, with few exceptions, the lesions found at necropsy were not typical croupous or lobar pneumonia. The acute fulminant cases which were seen in the spring of 1918 and more frequently in the fall epidemic were rarely present.

r Source of information, except as otherwise indicated: (1) Medical reports to the Surgeon General, 1917, 1918, and 1919; (2) Necropsy records, on file, Army Medical Museum, Washington, D. C.

63  CHART VII.- The incidence and fatality or the acute respiratory diseases at Camp Dodge


From September 20, 1917, until May 10, 1918, 675 patients whose cases were diagnosed as lobar pneumonia, were admitted to the base hospital.s   

The ordinary clinical lobar pneumonia, due to the pneumococus, prevailed until about March 20; then, abruptly the streptococcus type predominated, with a very great increase in the incidence of the disease. From September 20 until March 20, which marks the period of ordinary lobar pneumonia, 276 cases were treated.   

A very mild type of pneumonia prevailed during the early autumn, with a mortality in the first 100 cases of 7 percent. The onset was usually severe and evidence of intense toxemia prevailed, but the course of the disease was short. In the first 88 cases terminating by crisis, the duration in 7 was 2 days; in 11, 3 days; in 13, 4 days; in 13, 5 days; in 18, 6 days; in 9, 7 days; in 10, 8 days; in 2, 9 days; and in 5, 10 days or more. In this series of 276 cases, empyema was present in 31, or 11.2 percent. Even in these early cases, however, there was a marked tendency to multiple pus foci and the mortality from empyema was high. In those of this series where the type of pneumococeus was determined, Type I was found in 22.8 percent; Type II, typical and atypical combined, 46.8 percent; Type III, 7.6 percent, and Group IV, 22.8 per cent.   

The lobar pneumonia, presented the usual symtomatology. The only point of interest was the gradually increasing virulence, as manifested by the mortality from October 1, 1917, to January 1, 1918, and the gradually increasing frequency of empyema. In October this complication appeared in 2.3 percent of cases; in November, 2.2 percent; in December, 17.3 percent, reaching its maximum in January with 27.9 percent. Empyema was further characterized by loculated pus pockets, which accounted for the high mortality.   

No necropsy material from this period is available for study. As indicated in Chart VII, there was a moderate increase in respiratory disease during March and April, which gradually declined to its low point in August. The death rate shows a marked increase in March, continuing through April. Mild respiratory affections not requiring hospitalization, and therefore not appearing on the records of admissions, were very frequent.   

The epidemic of streptococcus pneumonia appeared suddenly between March 18 and 20, continued with great severity for six weeks, then gradually became less intense, still continuing, however, May 10 at the rate of four or five cases daily. The virulence of the epidemic, however, became less marked after the first three weeks, although the number of new cases remained high.   

Evidence of severe intoxication appeared very early; empyenia became very frequent and developed extremely early, two patients entering the hospital with pleural exudate who had been drilling the previous day. While the involvement in the lung maintained a lobar type, clinical evidence of complete consolidation was far from constant. Dullness with suppressed breathing and subcrepitant r les, but inconstant or localized bronchial breathing, were the usual findings; rusty sputum, provided there was expectoration, was the rule. Early roentgenoscopy showed that the infiltration was lobar in character. The devel-

The following statements of fact are hased, in the main, on: (1) Epidemic of Streptococcus Pneumonia and Enmpyenma at Camp Dodge, Iowa, by Joseph 1L. Miller and Frank B. Lusk, Journal of the American Medical Association, Chicago, l918, lxxi, No. , 702-704. (2) Pneumonia and Empyema at Camp Dodge, Iowa, by Joseph L. Miller: in Contributions to Medical and Biological Research, 1919, Vol. II. 1134-1137.


opment of an exudate was often exceedingly difficult to determine by the ordinary physical findings, aided by the roentgen ray, and it became necessary to resort to frequent exploratory aspirations. Often these were repeated several times before the fluid could be located. Early this exudate was only moderately turbid, contained numerous polynuelear leucocytes, and showed on smear short chain streptococci. Gradually the fluid became definitely purulent.   

The bacteriologic findings in 95 of these exudates showed pure streptococci in 88, all being hemolytic. Pneumococci combined with streptococci were found in three; two of Type II and one of Type I. Pneumococci without streptococei, but often combined with other bacteria, were found in four cases; one of Type I, two of Type II and one of Group IV.
In 596 cases of both types, the first involvement occurred in the lower left lobe in 253; lower right, 252; upper right, 45; upper left, 23; middle, 12; entire right lung, 5; entire left lung, 1; and in both lungs, 5.   

Complication with pus formation was very frequent. While empyema developed in 11 percent of the cases of pneumonia previous to the epidemic, it appeared in 34.8 percent during the epidemic. One case of suppurative peritonitis was found in the 276 cases of pneumonia preceding the epidemic and five cases in the first 364 streptococcus cases. In three of these it was associated with suppurative pericarditis and in all cases with empyema. It was interesting that four of these cases developed within a few days of each other. Suppurative pericarditis was found in 15 of the 142 cases of empyema, or 10.5 percent. It was found only in association with empyema, and in only one case was it found in a colored soldier. Rectus abscess, which was more or less frequent in similar epidemics in other camps, was not present in the first 364 cases, but one abscess developed in a more recent case. The colored soldiers developed empyema in 20 percent of the pneumonia cases and the white soldiers in 45 percent. The empyema mortality was 44 percent in the colored and 64.8 percent in the white soldiers. This lessened tendency to empyema and suppurative pericarditis in the colored troops was striking. The deaths from uncomplicated streptococcus pneumonia in the colored was 19.97 percent, and 10.7 percent in the white.   

The mortality in 142 cases of streptococcus empyema was 60.4 percent. This is considerably higher than the average in other camps. There is no apparent explanation of this unusually high mortality, as the method of treatment was simiilar to that carried out elsewhere. It is true that in only a moderate percentage of these cases, death was due not to empyema itself, but rather to severe infection, the majority of the patients dying with merely a turbid fluid, the time being too short for it to become thick and creamy.   

The cases of streptococcus empyema may be divided into three groups: (1) Those who die early, no matter what form of treatment is instituted, from acute toxemia; (2) those with multiple pus foci, difficult to detect, because of the inability to locate and drain all foci; these all die, and (3) those who usually recover, either from early operation or aspiration followed by operation; here are included those with moderate toxemia and those with localized pus accessible to drainage.


The mortality in 364 streptococcus pneumonias was 32.5 percent as compared with a mortality of 11 per cent in the previous pneumococcus type.   

Arthritis as a complication appeared in only six cases. Suppurative otitis media was quite common. Erysipelas developed as a complication in seven cases. These were facial except in one case, where it developed at the site of the exploratory puncture.  

The pathology of this acute outbreak presented certain interesting aspects.t   

The organism found in the great majority of the cases was the streptococcus hemolyticus. It produced its initial infection in the oral cavity in the form of pharyngitis. Sometimes it gave rise to tonsillitis, rarely of the follicular type; often an exudative membranous inflammation extended over the tonsils, the uvula, and the pharynx. In this was found practically a pure culture of this organism. There was produced, then, first of all, a pharyngitis, which was likely to last for weeks, the patient then becoming a true carrier. If he had good resistance, the process stopped at this stage, and this was usually the case. However, the infection might extend to the accessory sinuses of the nose, it might enter the frontal sinus, it might produce otitis media, and, in some cases, a streptococcic meningitis. The route which the organism usually took, in a susceptible individual, was downward through the pharynx to the trachea, and into the bronchioles. Sometimes it produced only an ordinary catarrhal inflammation, but very frequently the infection was more severe than that, so that there developed a hemorrhagic tracheitis and laryngitis, with necrosis, at times, of the bronchi and of the arytenoid and thyroid cartilages. The process might stop, however, before extending into the bronchioles and the person entirely recover. Or it might go on, producing a bronchiolitis, which often was suppurative, with extension of the process into the lung substance.   

The infection, then, apparently extended out through the bronchial walls. These became attacked and infiltrated with lymphocytes, following which the infection spread peripherally into the lung substance, so that a number of nodules simulating, to some extent at least, tubercles, developed in the lung. These usually spread, attaining a size as great as 1 centimeter in diameter. The lungs were studded with these nodules. The infection extended through the lymphatics to a great extent. The nodules then became confluent in some cases, producing clinical lobar pneumonia, which was not a true lobar pneumonia, but bronchial pneumonia, in which most of the lung substance between the nodules had become involved. The area involved, as a rule, was not so firm and did not contain the plugs of fibrin which one sees in the ordinary pneumonia due to the pneumococcus. In other cases these nodules occurred in patches, a true lobular pneumonia which is also due to the streptococcus. In some other cases the streptococcus infection followed a true pneumococcus pneumonia.   

The empyemata formed were often very peculiar, differing greatly from empyemata usually seen. The fluid rarely filled the entire pleural cavity, but at times a large amount was present. Sometimes the fluid was thin and but slightly cloudy due to the presence of but few leucocytes. The most striking feature was the marked tendency toward encapsulation, often forming a number

t The following discussion is based on: The Pathology of Streptococcus Infection of the Lungs, by D. J. Glomset, Journal Iowa State Medical Society, Clinton, 1919, ix, No. 4. 143-145.


of pockets. These were made by the deposits of fibrin. In many cases the deposits on the parietal pleura were over a centimeter thick. This fibrin sealed off the fluid-seeping areas, forming encapsulating empyeinas, which were found in the most unlooked-for places. Occasionally there was found an apical empyema, encapsulated so as to cover half of the upper lobe. Again, there were found various pockets, some smaller, some larger, under the sternum, in the mediastinum, and especially between the lobes, and more particularly between the lower lobe and the diaphragm.   
The next complication in order of frequency was pericarditis, of which there were many varieties, from cases in which only a few floccules of fibrin were present with a slight increase of the fluid, to cases in which the sac was distended with a thick, greenish pus.   

In several cases the mediastinum was absolutely filled with purulent material. Sometimes localized pockets of pus were found between the esophagus and the trachea, and in two cases at least death was due to pressure on the heart.   

The next most frequent complication was an inflammation of the peritoneum. This occurred in 10 cases. The peritonitis was generalized and not so much fibrin seemed to be formed there as in the thorax.  
The high incidence of fatal pneumonia at this camp caused the Surgeon General to send specially qualified medical officers to Camp Dodge to study the condition. As a result of the studies there and at other camps, attention of the medical profession was directed to the condition described as "interstitial pneumonia " and attributed to the streptococcus hemolyticus, which was found practically constantly.u   

Usually a few days after the onset of measles, or a few days after convalescence from that disease, the patient had pain in the chest, sore throat, cough, a chill, high fever, and general malaise. The physical signs were indefinite, but there was usually some vague dullness over the chest and rales were scattered widely throughout both lungs, especially at the bases and behind. The patient became dyspneic, the respiration being especially labored in inspiration, cyanosis of a curious livid type appeared, and there was a great deal of nervous excitability and apprehension. The dyspnea produced extreme discomfort and the most violent inspiratory efforts were made, bringing into play all the necessary muscles; this made sleep impossible. The eyes were bright and shining, the mind most alert and filled with dread.   

After a few days, dullness or flatness over the thorax became apparent and an exploratory aspiration revealed the presence of turbid fluid in the pleural cavity. The accumulation of this took place with extreme rapidity, and when removed it was replaced in a surprisingly short time.   
Delirium was not uncommon, and death often occurred after 10 days or 2 weeks of such illness, although in some cases the whole course of the disease from the onset of the bronchial symptoms to death was only four or six days. Hoarseness or complete aphonia was especially characteristic of the cases which developed most rapidly and intensely after measles, but it sometimes occurred in those who had not had that disease.

The Pathology of the Streptococcal Pneumonias of the Army Camps, by W. G. MacCallum; in Medical Clinics of North America, Philadelphia, September, 1918, 379-391.


At autopsy in the cases in which death occurred a few days after the onset of the disease the pleural cavities often were found to contain no excess of fluid and their surfaces were smooth and glistening, In the great majolity of the cases, however, there was present a considerable amount of exudate in at least one pleural cavity, sometimes in both. The pleural surfaces were sprinkled with small petechial hemorrhages, especially over the back of the lung, and covered with a thin rough exudate of fibrin. In such instances the fluid was thin and watery, turbid and brown or greenish brown, with a granular sediment easily stirred up from the bottom and many floating shreads or flakes of fibrin. It was swarming with the streptococci in chains usually unmixed with any other organism. In cases which ran a longer course the fibrin on the surface was thick, yellow, and shaggy. There were often to be found adhesions between the lobes, between the lung and the costal pleura or diaphragm, or in such an arrangement as to enclose separately a space between the mesial surface of the lung, the pericardium, and the diaphragm.   

There the fluid exudate was distinctly purulent in character and either opaque greenish yellow or brownish gray with a slight bloodstaining. The disposition of the adhesions often allowed of the encapsulation of pockets of this pus between the lobes or elsewhere, especially often in the situation above described between the lung and the pericardium. Incision into the surface of the lung showed that the pleura itself had become greatly thickened by being converted into a layer of granulation tissue which tended to organize and replace the fibrinous exudate. The rapidity with which this organization occurred was surprising. When the pleura had been evacuated by operation the character of the exudate and the thick layer of granulation tissue upon the surface was unchanged, unless secondary infection had occurred, when the pus might become exceedingly foul.   

The effect of this great accumulation of exudate in the pleura, which might amount to 1 or 2 liters or more, was to cause the extensive collapse of the lung, which appeared as a blue pasty mass of tissue plastered against the mediastinal tissue. This was not the only cause of collapse, however, since obstruction of the bronchi played a part, and it frequently appeared in parts of the opposite lung even when there was no exudate in that pleural cavity.   

The lung, in general, was flabby, but there might be felt throughout its substance nodules or larger areas of consolidation. The bronchi exuded a thick brownish purulent exudate, and the bronchial glands were enlarged and soft. One section of the lung itself presented in different cases a very great variety of appearances. In some there was the anatomic complex of interstitial bronchopneumonia. in others, lobular pneumonia, and it must be remembered that these were very frequently combined.   

In the early stages of interstitial bronchopneumonia there was already a patchy atelectasis of the lung, the bronchi were filled even to their smallest branches with opaque yellow pus and were surrounded by a halo of hemorrhage. Otherwise the lung substance was air-containing in part and in part edematous. In all cases miscroscpic examination showed the bronchioles filled with polymorphonuclear leucocytes, and streptococei often in tangled masses ranged along


the wall. The lymphoid tissue in the bronchial wall was slightly swollen through an increase in the number of its lymphoid cells. The adjacent alveoli were filled with blood and contained a few leucocytes.   

Somewhat older cases showed on section numerous firm projecting nodules throughout the lung, which on close inspection were found to be cross-sections of bronchioles with surrounding consolidation. If the bronchioles happened to be cut longitudinally, these condensations of the tissue appeared not as nodules, but as very thick-walled tubes further surrounded by a mantle of consolidated lung tissue. The area to which such a bronchus, obstructed as it was with a mass of purulent exudate, should supply air was collapsed completely. The consolidation was scarcely in this area, but rather in the alveoli adjacent to the bronchus. These, too, were sometimes surrounded by hemorrhage, but usually that was only in the freshest cases. Viewed on cross section the tiny bronchiole, with its thickened wall and surrounding area of consolidation, projected like a tubercle from the cut surface. Indeed, it is probable that when these were fairly small and uniform they were often mistaken for miliary tubercles. As they increased in size they became more and more nearly confluent, and the intervening tissue showed a viscid edema which drove out its original content of air. In time, quite extensive areas showed a patchy, irregular consolidation through the coalescence of such peribronchial areas.   

Microscopically the bronchial wall was still found to be the part most intensely affected; the streptococci were present in the purulent exudate in about the same numbers. Much of the epithelium was found to be desquamated and in some cases of most intense infection the whole lining of the bronchus was necrotic and coagulated, and appeared as a diphtheritic pseudomembrane. The wall of the bronchus was greatly thickened and densely infiltrated with mononuclear cells of the type of lymphoid cells, with some larger forms. It is probable that much of this represented a hyperplasia of the lvmphoid tissue normally found scattered in the wall and condensed at the angles of division. This was especially indicated by the peculiar course of the blood vessels in this cellular tissue, which was that of the blood vessels of the lymph nodes. But much of the thickening of the wall was due to actual new formation of connective tissue cells, to edema and hyperemia, and to a true infiltration of wandering mononuelear cells. It assumed the appearance in many cases of a richly vascular granulation tissue in which the new blood vessels were radially arranged, and in which, in the case of larger branches, remnants of mucous glands were to be found. The infiltration with mononuclear cells extended to the walls of the adjacent alveoli, which were thereby much thickened. These alveoli contained a dense exudate which was no longer composed of blood, but of desquamated epithelial cells, mononutclear cells, and solid plugs of fibrin. Further out the fibrin was replaced by fluid which appeared to be especially thick and viscid. No streptococci were discoverable in the substance of the bronchial wall or alveolar walls, or even in the exudate in the alveoli. They were present, however, in abundance in the lymphatic canals in the bronchial walls, and there they were entangled in a thrombus mass which included also many mononuclear cells. Such lymphatics extended in this thrombosed condition throughout the lung and communicated, on the one hand, with the sinuses of the lymph-


nodes at the hilum of the lung, on the other, with the network of lymphatics in the pleura. It was thought that it was through them that infection extended from the bronchi to the pleura. This was against the natural current of the lymph, but the organisms extended by growth along the obstructed lymphatic canals.   

The blood vessels were not thrombosed or otherwise altered except that their adventitious tissue was densely infiltrated with cells. This, again, might have been partly due to hyperplasia of the normal lymphoid tissue, especially since such infiltration after extending outward from the bronchial walls faded away in the alveolar walls until one approached blood vessels or interlobular septa, when it increased again.   

The interlobular septa were greatly widened and most conspicuous on the cut surface of the lung. This was due, first, to their infiltration with fluid cells and fibrin, which later became replaced by connective tissue, and second, to the presence in their course of lymphatic canals which were thrombosed and distended with opaque yellowish white material. Such thrombosed lymphatics stood out on the cut surface of the lung running to the pleural network as huge opaque beaded strands sometimes large enough to be mistaken for bronchi.   

Finally the organization of the exudate in the alveoli and in the bronchi and its replacement by strands and columns of connective tissue ultimately clothed in alveolar epithelium, formed a striking feature of the late stages of this process. The thrombosed lymphatics were similarly converted into solid vascularized cords of fibrous tissue, which indicated the need for a most extensive new formation of lymphatics if absorption from the pleura was to be resumed. Then, as described above, the pleura itself became a thick red velvety layer of granulation tissue through the organization of the overlying fibrin.   

The restricted distribution of the streptococci was most interesting. They were found in the bronchi, in the lymphatics, and in the pleural exudate, where they were limited practically to the surface layer of the fibrin and the fluid. The whole process indicated their inability to invade the tissue proper, and showed that the body was offering a strong resistance and erecting massive barricades against their entrance.   

All this was quite different in the second form of the pneumonic process, designated "lobular pneumonia." In these lungs patches of consolidation were found associated with wide zones of hemorrhage, but in no such regular relation to the bronchioles as has just been described. These patches might be terminal, occupying the area supplied by a bronchiole, or they might be confluent and large enough to occupy one or several lobules; sometimes they had the granular appearance of the cut surface of a lobular pneumonia, but more often they were elevated, dry, dull and opaque, and thickly surrounded by dense hemorrhage. There was no especial thickening or prominence of the bronchi. The interlobular septa were not necessarily prominent and often the pleura was but little thickened. The lymph channels, however, were thrombosed and very large and conspicuous in many cases. Microscopically there was no remarkable change in the bronchi except that they were full of purulent exudate and blood, but all the alveoli in the area were packed tightly with leucocytes in a network of fibrin, and among these leucocytes and often enclosed within them were


great quantities of streptococci uniformly scattered throughout the alveolar contents, a condition totally at variance with that found in the interstitial bronchopneunmonia. The alveolar capillaries were sometimes occluded by hyaline fibrinous thrombi. Whole areas of such consolidated and intensely infected lung became necrotic, and it was especially about such patches that extensive hemorrhage occurred. In these necrotic patches the streptococei grew out into long chains which formed a dense tangle.   

Small areas of such exudation into the alveoli with abundant streptococci were quite often found in the most intense and acutely developed lesions of interstitial bronchopneumonia constituting a combination of the two lesions. The explanation of the existence of two such distinct types of reaction to the invasion of the streptococcus is not entirely easy. It seems very similar to the condition in tuberculous infection of the lung. In one case there may be a few tubercle bacilli setting up the formation of tubercles, tuberculous invasion of the lymph channels, interstitial induration, etc., while in another case, or later or earlier in the same case, there is a sudden exudation into a whole area of pulmonary tissue, and the production of a gelatinous and caseous tuberculous pneumonia with great numbers of bacilli and rapid destruction of the tissue. It must be due to a remarkable difference in the power of resistance to invasion on the part of the tissues. It has been suggested that this sudden invasion of the tissue by bacteria which were formerly held within bounds is due to a sensitization of the tissue by the previous infection. Possibly the same explanation might hold in this case, although the whole process occupied a very short time. At any rate, that seems to be another way of expressing the idea that resistance in one case is strong, in the other almost absent.   

In the cases of lobular pneumonia great areas of necrotic tissue often became coagulated, opaque, and yellow, and appeared as abscesses which might occupy a large part of a lobe. This was not limited to the obvious cases of lobular pneumonia, but sometimes occurred in those in which interstitial changes were prominent. The liquefaction of this material and its discharge through a bronchus left a ragged cavity which might in some instances, through forming a communication with the pleura, set up a fistulous connection between the pleura and a bronchus.   

To summarize, it may be said that an extensive epidemic of pulmonary disease resulted from widespread infection with a hernolytic streptococculs which invaded most readily upon the bases of a predisposing attack of measles. This produced inflammatory and ulcerative changes in pharynx and larynx and,extending to the bronichioles, set up a bronchopneumnonia which, in persons of poor resistance, might take the form of a lobular pneumonia with extensive spread of the organisms and rapid necrosis, or in those with better powers of resistance it might produce a purulent bronchitis with great thickening of the bronchial walls and consolidation of the adjacent pulmonary tissue, infection of the lymphatics, and thence of the pleura. Induration of the framework of the lung and organization of the exudate tended to limit the process. In both types pleurisy with abundant effusion was very common.   

Uncomplicated influenza was diagnosed in a large number of instances at this camp prior to the pandemic, with a peak of 939 cases in the month of March,


common respiratory diseases for this month being 1213. The diagnosis of influenza was made in 136 cases in July, 28 in August, 573 in Septeniber, 6,480 in October, 1917.   

During the week preceding the sudden appearance of the epidemic of September 28, 1918, three distinct outbreaks of an infectious nature occurred in widely separated sections of the base hospital at Camp Dodge.v Pharyngeal cultures from these cases showed Streptococcus hemolyticus and Bacillus influenzae present in unusually large numbers. Admissions to hospital during this period increased moderately in number, and an unmistakable but not alarming number of acute nasorespiratory disturbances, not unlike similar clinical conditions of the preceding month, gave warning of impending trouble.   

September 29, 1918, laboratory workers detailed to the admitting office, cultured and made white blood counts on 152 cases. It was found that the Bacillus influenzae was present in 66 percent, Streptococcus hemolyticus in 45 per cent, and the M. catarrhalis in 71 percent of the cultures, and that the average white cell count was 9,300.   

Further observation of these cases revealed that 890 white blood counts ranged as follows per day: First day, 9,300; second day, 8,000; third day, 8,200; fourth day, 6,000; fifth day, 7,200; sixth day, 7,700; seventh day 8,200; eighth day, 8,200.   

Between October 3 and October 11, 53 complete autopsies were performed. The following data were compiled from the records on the first 53 cases of epidemic pneumonia autopsied.   
A purplish mottling of face, neck, and dependent portions was present in nearly all cases dying within two weeks of onset. The gross pathological lung lesions divided the findings into three classes or types of pneumonia: he broncho-interstitial type, which showed small interstitial hemorrhagic areas about moderately thickened bronchi of varying size, standing out prominently from the sectioned surface and exuding dark red blood. The lobular type, in which hemorrhagic areas of infiltration of varying sizes, without noticeable bronchial thickening, were encountered. Both types were not unlike the conditions found and described by MacCallum during the pneumonia epidemic of May, 1918 (q. v. supra). Several cases were found having in one lobe lesions characteristic of the broncho-interstitial type, while the adjoining lobe revealed a picture of the lobular form. To this condition the term mixed or broncho-lobular pneumonia was given. This nomenclature, while describing accurately many of the lung lesions grossly should not be relied upon too implicitly, as further histological study may place a much larger number of cases in the mixed or broncholobular class.   

The number and percentages of the various types of pneumonia and enpyema found at autopsy in 53 cases were as follows: Broncho-interstitial, 23; lobular, 18; mixed, 10; lobar, 2. The empyemas, 22 in number, were bilateral in 7, right in 5, left in 10. In 19 cases, infected serosnaguineous pleuritis was present with lobular pneumonia in 12, broncho-interstitial in 4, mixed in 3.


vThe following statements of fact are based in the main on: Laboratory Report on Epidemic Pneumonia, Camp Dodge, Iowa, by William G. Dwinnell, American Journal of the Medical Sciences, Philadelphia, 1919, clviii, n. s., 216-232.


Of the 31 cases dying without empyema, 29 died within 2 weeks of onset of the first symptoms, the average duration being 7 days; the extremes were 3 to 13 days. The 2 dying later than 14 days died at 15 and 44 days, respectively, from the date of onset.   

Because of the uncertainty attending the onset of the pneumonia the duration of the disease in all cases was figured from date of onset of first symtoms of illness.   
The important complications in 53 cases were acute adrenalitis 28, mediastinitis 19, pericarditis 9, peritonitis 6, necrosis of the lung 4.   

The Streptococcus hemolyticus was recovered in 80.7 percent of the cases from the lung, the pleural cavity, or the heart's blood, and from the last source in 52 percent of the cultures. The Bacillus influenae was recovered in five cases, or 9.6 percent of the cultures, and was found in three cases in both the lungs and pleural cavities, in the heart's blood once, and in the peritoneal cavity once. The Bacillus influenzae was found associated with the Streptococcus hemolyticus four times and the Staphylococcus aureus once. A non-hemolytic streptococcus was recovered five times and the Staphylococcus aureus twice.   

The low incidence of Bacillus influenzae found was undoubtedly due to poor technique in handling the cultures. The large number of positive cultures obtained from the 69 cases of the second series mav be ascribed to excision of lung tissue and direct smearing of freshly-made human blood-agar plates adapted to influenza work at a reaction of 0.2 acid of phenolphthalein and containing two drops of blood to 200 c. c. of glucose agar, meat infusion. This medium gave more positive cultures than the laked rabbit blood previously used.   

The cultures reported as Bacillus influenzae did not grow as typical dewdrop colonies as described by Pfeiffer and others. The colonies were less definitely outlined, slightly granular in the center, suggesting a faint opaquish color. There was no growth on plain agar. The bacilli varied in size from a coccobacillus to a rod of slightly larger than the typical Pfeiffer bacillus in the original culture. Mannite, sucrose and lactose were not fermented. One cubic centimeter of moderately heavy emulsion was not lethal to either mice or guinea pigs.   

No post-mortem work was done between October 11 and 23; 69 necropsies were made between October 23 and December 1, 1918. The necropsy findings were as follows: Pneumonia, broncho-interstitial, 26; lobular, 27; mixed, 14; lobar, 2. Enipyema, bilateral, 21; right, 19; left, 8. Serosanguineous pleuritis was bilateral in no case, right in 2, left in 1.   
Of the 21 cases dying without empyema, 11 died within 2 weeks from the onset of the first symptoms, the average duration being 7 days, the extremes being 1 to 12 days; the 10 dying later than 14 days averaged 24 days; extremes, 5 to 54 days. Eleven of the 48 empyema cases lived an average of 11 days, extremes 5 to 14 days. The remaining 37 cases averaged 32 days; extremes, 17 to 72 days.   

The important complications that occurred in the 59 cases were: Acute adrenalitis, 45; pericarditis, 20; mediastinitis, 19; necrosis of lung, 17; pleural pockets of pus, 17; peritonitis, 6; interlobular pockets of pus, 11.


The bacteriology in the 69 cases was as follows: Hemolytic streptococcus, 41; Bacillus influenzae, 34; nonhemolytic streptococcus, 17; Type I pneumococcus, 1; Type II pneumococcus, 2; Type III pneumococcus, 4; Group IV pneumococcus, 1.   

Mediastinitis occurred 38 times in 122 cases. It was associated 32 times with empyema. Three cases, showing general subcutaneous emphysema of the face and neck to Poupart's ligaments, were associated with mediastinitis.   

Pericarditis occurred 29 times, in 15 of which mediastinitis was present, and in 22 of which empyema was present. There were 12 cases of peritonitis, 9 of which were with empyema, 1 with lung abscess, and 2 were associated with operative procedures. Only two of the four cases of lobar pneumonia showed no other type in the lungs.   

The predominant organism in this camp during the entire period of its existence was the hemolytic streptococcus. Lobar pneumonia due to the fixed types of pneumococei occurred in considerable numbers during the fall of 1917 and the following winter and spring. In the influenza period the access of the streptococcus to the injured lung resulted in fewer instances of the hemorrhagic type of pulmonary inflammation discoverable at necropsy.   

From the reports cited above, it is seen that, with the exception of the period of greatest intensity of the influenza pandemic, when the exigencies of the situation prevented it, the pathology of the respiratory diseases at this camp were unusually well studied and reported. It appears, however, that, as occurred in most camps, no line could be drawn between the pathology of clinical influenza and of the secondary infections which followed or occurred practically simultaneously. It is not at all certain, for instance, that some of the lesions attributed to the streptococcus, in which that organism was scarce or absent from considerable areas microscopically, were in fact due to that organism. It is quite possible that such lesions as the acute reaction about terminal bronchioles and the hemorrhagic lesions occupying more or less the whole of lobules were due to the bacillus of Pfeiffer or to the cause of influenza. Cultural work during the spring of 1918 and even early in the pandemic period was admittedly inadequate to show the incidence of B. influenzae even in the material cultured.   

Streptococci of the hemolytic variety appear to have been responsible for the high case fatality of all respiratory diseases, and were associated with practically all types of pulmonary lesion, including lobar pneumonia, though in many cases it was evident that the invasion by these organisms was secondary not only to the virmus of measles or influenza but also to pneumococcal lesions, the majority of which followed measles or influenza.
The respiratory diseases of this camp are represented in the collections of the Army Medical Museum by 232 protocols of necropsies and 59 lung specimens with some of the other tissues from the necropsies. Suitable stains demonstrate numerous Gram-negative, minute coccobacilli in many of the well fixed tissues from cases in which no such organism was isolated in culture, and organisms of the pneumococcus type, as well as streptococci are found in the pulmonary parenchyma from cases in which only streptococci were isolated in culture.


Study of the histopathology of this camp yields convincing evidence that bacteriologic methods considered of the highest grade at the time were not to be depended on to reveal all the organisms present in the pulmonary inflammations of this period.   

Study of the cases diagnosed interstitial pneumonia reveals the fact that in the majority of instances the perivascular lymphatics are as much involved as the peribronchial, and in some instances the oldest or farthest advanced process is along the vessels rather than the bronchi. This is strong evidence that the streptococcal invasion is a spreading lymphangitis in such cases, with abscess formation and secondary invasion of the bronchial lumen. The pneu- monic areas in these cases are of two types: First, that extending a varying distance into the lung from the purulent lymphatics, and second, nodular areas, representing the parenchyma supplied by a smaller bronchiole, which resemble and may well be pneumonias of the inhalation type resulting from the rupture of abscesses or dilated lymphatics into the bronchial lumen, just as happens in tuberculosis, to which MacCallum called attention as having points of resemblance in the varying types of tissue reaction and bacterial content of the lesions.


Mobilization at this camp did not start until the first of the year 1918, and as the first increments were seasoned troops, respiratory diseases in epidemic proportions did not occur until late. The epidemic rise, with peak in April, was due to influenza and common respiratory diseases, the case fatality being low. Case fatality for all respiratory diseases increased prior to the onset of the influenza epidemic.x   

The expected storm of the prevailing pandemic infection broke suddenly on the 8th Division, at this camp, October 8, 1918, and during the next 6 weeks 2,418 patients suffering from respiratory diseases were admitted to the base hospital. In addition, many soldiers having more or less mild infections were cared for in the various camp infirmaries in order to avoid overcrowding in the base hospital. Altogether there were, at a conservative estimate, 3,000 cases. Pneumonia was diagnosed in 408 cases, an incidence of nearly 14 percent. We know now, however, that there were many cases of pneumonia that were diagnosed as bronchitis, and that the true incidence of pneumonia was greater than that indicated.   

Of the 408 patients with pneumonia diagnosed, 147 died, a mortality of 36 percent for the pneumonia series and about 5 percent for the epidemic. No deaths occurred without a complication of either lobar pneumonia or bronchopneumonia.   

Epistaxis was a common feature throughout. At first it was looked on as merely an incidental occurrence, as in typhoid fever, and not as possessing special significance. Instances of epistaxis multiplied, however, and often blood was seen to gush from a patient's nose and mouth. When pneumonia appeared, the pa-


w Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1917, 1918, and 1919.

x The following statements of fact are based, in the main, on: Pandemic "Influenza" and Secondary Pneumonia at Camp Fremont, Calif., by Walter V. Brem, George E. Bolling, and Ervin J. Casper. Journal of the American Medical Association, Chicago, 1918, lxxi, No. 26, 2138-44.

76   CHART VIII.- The incidence and fatality of the acute respiratory diseases at Camp Fremont


tients often spat quantities of almost pure blood. Female patients, of whom there were altogether about 100, had a hemorrhagic vaginal discharge, which at first was thought to be coincident menstruation, but later was interpreted as hemorrhage from the uterine mucosa. Finally, the extraordinary hemorrhagic picture seen at necropsy, together with the constant leucopenia in the initial stages, completed the evidence that led to the belief that an important feature of the disease was some kind of blood dyscrasia, which was of grave significance and suggestive of purpura hemorrhagica. In a few patients, large purpuric areas developed in the skin.  
It was believed that this condition was due to a depressant influence exerted bv the initial infection on the bone marrow, and that it was closely related to the next extraordinary feature of the disease.   

In the influenzal stage, leucopenia was practically always found, and at first it appeared to be a feature also of the pneumonic stage. Indeed, one saw many full-blown cases of apparently lobar pneumonia with a leucocyte count of less than 5,000; there were five cases with a count of 3,000 or less. Smears from the lungs of some of the fatal cases showed innumerable diplococci with the large pneumococcus type of capsule, and the cultures were positive for pneumoccoci.   

But other pneumonia patients with early leucopenia recovered, and when subsequent counts were made, it was found that in each case there was a gradual rise to from 10,000 to 24,000 leucocytes.   

The conviction gradually developed, therefore, that the influence of the initial infection was a depressant one on the bone marrow resulting in leucopenia, while, on the contrary, the pneumonic infection tended to stimulate, as usual, the production of leucocytes, which gradually increased in number in the blood stream as the depressant influence waned.   

Two characteristic types of pneumonia were differentiated. The first, and most common type, was that which, when well developed, gave the physical signs of lobar pneumonia. The onset usually occurred when defervescence of the symptoms of the initial infection was almost or quite complete, about the third or fourth day. It was ushered in frequently by a chill, a sharp rise of temperature, a localized explosive shower of fine rȃles usually about the angle of the scapula and more frequently on the left side, and bloody sputum, amounting often to marked hemoptysis. A few roentgenographic examinations at this stage showed shadows indicating that already consolidation was beginning. Occasionally the milder of these cases cleared up quickly without further signs of pneumonia developing, but most of them proceeded to a stage of definite signs of consolidation of one or more lobes.   

The temperature curve was irregular, frequently not rising as high as from the initial infection, and usually ending by lysis, the mean duration of the pneumonia fever being about seven days. The pulse rate was slow in proportion to the fever, and the blood pressure was low, the systolic pressure frequently being 100, and the diastolic pressure from 40 to 50. The pulse pressure was good, and there were no signs of cardiac insufficiency except in the fatal cases after the onset of respiratory insufficiency and shortly before the end. The danger never appeared to arise from the effect of the infection on the cardiovascular mechanism.


The respiratory rate, likewise, was slow in proportion to the fever until the toxemia became marked, when the respirations became accelerated and the respiratory abdominal excursions limited. The chief toxic effect of the infection seemed to be exerted on the respiratory mechanism, failure of which, due to excessive toxemia and to the mechanical effect of extensive consolidations and hemorrhagic edema of the lungs, caused the death of patients.   

It was not known whether this type was a true lobar pneumonia or a massive bronchopneumonia due to the fusion of small areas, but it was believed that it was a lobar pneumonia with the pathologic picture modified by the hemorrhagic exudate and the paucity of leucocytes.   

The lungs at necropsy showed massive, air-free consolidations without any definite lobular pattern; no pus could be expressed from the bronchi; the bronchial mucosa did not show a purulent exudate or any special inflammatory thickening or change other than narked congestion, and there was practically no fibrin in the exudate. These findings were in marked contrast to the true bronchopneumonia type.   

The second type was that of true bronchopneumonia. The onset was not clearly defined, the course being continuous with that of the initial infection; the temperature curve was irregular; the sputum was not bloody, but purulent (nummular); physical signs of diffuse bronchitis were usually present. Fatal cases sometimes showed a low or even normal temperature with cyanosis and marked "air hunger." The patient belonging to this group had a normal temperature, good pulse, no sputum, and no r les in the chest. The percussion note over the lungs was hyperresonant, the breath sounds distant and scarcely audible; there was marked dyspnea and cyanosis. The diagnosis was not suspected until necropsy disclosed the real condition. The bronchi were full of pus; and little nodules, bronchopneumonic consolidations, were scattered uniformly throughout the lungs. Death appeared to be due to actual asphyxiation..   
Five patients with this type of pneumonia developed a puzzling subcutaneous emphysema, which, in the most marked case, extended from the forehead to Poupart's ligament in front and the buttocks behind. The post-mortem examination demonstrated in two cases that emphysematous air cells at the root of the right lung had ruptured into the loose mediastinal tissues, and that the air had progressed upward to the neck, and infiltrated the subcutaneous tissues. Cultures from the subcutaneous tissue for anaerobic organisms were negative.  
The clinical features of this type, and also the pathologic features, to be described, indicated that it was due to a direct extension of the inflammation down the bronchial mucosa. B. influenzae was found in cultures from the pus droplets of the small consolidated areas in the three cases carefully studied bacteriologically at necropsy-twice in pure culture, one with M. catarrhalis.   

The leucocyte counts in this group were few. In two fatal cases there were 7,600 and 6,000 leucocytes, respectively. In the second case the count lose to, 16,400 before death. B. influenzae was present in pure culture in the lungs in both cases.   

The relative percentages of the two types of pneumonia could not be estimated, for they were differentiated but late in the epidemic, and the clinical records did not enable one to separate them. It is certain, however, that the first type was far more common, but the second type was by no means infrequent.


Deaths were always due to pneumonia. In 30 post-mortem examinations, lobar pneumonia was diagnosed 25 times, bronchopneumonia, 5 times. The two types presented striking differences at the post-mortem examination.   

In the lobar pneumonia type, frothy, bloody serum poured from the nose and mouth when the body was moved or the head lowered. On section of the skin, dark blood oozed from the cut vessels. The pleural cavities always contained a quantity of serosanguineous fluid, but the pleurae did not show the fibrinopurulent exudate commonly found in lobar pneumonia. The lungs were extensively involved, there being practically always large areas of consolidation in several lobes; hemorrhages under the pleurae were frequently seen; the lungs were dark red and irregularly mottled, very moist and bloody, the cut surface smooth and velvety; the consolidated areas contained no air and no pus could be expressed from the bronchi. The early condition was that of acute hemorrhagic edema. The peribronchial lymphatic glands were very large, dark red, congested, soft and friable. The heart muscle was dark red, relaxed and flabby, offering a strong contrast to the firm, contracted left ventricle nearly always present post mortem in bodies of patients dying from lobar pneumonia; the heart's blood was fluid and dark, there being only small, flimsy, post-mortem clots and no "chicken fat" clots. The abdominal viscera showed marked con- gestion, the glomeruli of the kidneys standing out distinctly. Microscopically, the alveoli of the lungs were full of red blood corpuscles in the portions recently involved; in the older portions polymorphonuclear leucocytes became more or less prominent in the exudate; no fibrin was seen; the bronchi were not plugged with pus; the blood vessels were engorged.   

In the bronchopneumonia type the tissues of the bodies were dry; there was no "frothing at the mouth;" no blood oozed from small cut vessels; the pleural cavities contained no fluid, but there were patchy areas of fibrinopurulent exudate with recent, easily separated plural adhesions. The lungs were very pale and extremely voluminous, and did not collapse when the chest was opened. Emphysematous bullous-like areas beneath and elevating the pleurae were common; these had transparent bubble-thin walls and were divided into numerous compartments by the walls of the dilated air cells. Pus streamed from the trachea when the lungs were removed. Small areas, from 2 to 5 mm. in diameter, of consolidated tissue were scattered uniformly throughout the lungs, and these little areas felt like disseminated conglomerate tubercles. On section, the lungs were dry, the solid areas were gray and granular in appearance, and a drop of pus could be expressed from the center of each. In one case, about half of the right upper lobe was consolidated; it had a mottled gray, granular appearance, and there were numerous points where droplets of pus could be expressed, the large area of consolidation being evidently formed by the confluence of smaller areas of bronchopneumonia. The peribronchial lymphatic glands were very large, the cut surface gray and granulated. They were soft and friable. The heart showed nothing remarkable, nor did the abdominal viscera. In one case the spleen was quite large and the malpighian bodies stood out distinctly, and the liver and kidneys showed cloudy swelling. In two necropsies the mediastinal and subcutaneous tissues showed marked emphysema due to the rupture of dilated thin walled air cells on the anterior surface of the


root of the right lung into the loose mediastinal tissue. Microscopically, the consolidated areas of the lungs showed a bronchiole full of pus cells and the surrounding alveoli packed with pus cells and desquamated epithelium, and there was considerable fibrin formation. The air cells of the adjacent uninvolved lung were dilated and the walls were thin; the vessels were not engorged.   

Cultures from the nasopharynx were made from every patient admitted to two of the wards; 148 patients were examined, and B. influenzae was isolated from 37 patients, or 25 per cent. Ten of these patients, or 27 per cent, developed pneumonia.   

Similar cultures from the nasopharynx were made in 537 cases. These were usually selected severe cases of respiratory infections.   

The results were as follows: B influenzae, usually associated with other organisms in 259 cases, or 46 percent; pneumococci in 149 cases, or 28 percent; streptococci, all kinds, in 160 cases, or 30 percent; staphylococci, all kinds, in 64 cases, or 12 percent; and M. catarrhalis in 62 cases, or 11.5 percent. Cultures from the nasopharynx and sputum of 158 pneumonia cases showed B. influenzae in association with other organisms 58 cases, or 38 percent; pneumococci in 64 cases, or 40.5 percent; streptococci in 46 cases, or 29 percent; staphylococci in 21 cases, or 13 percent, and M. catarrhalis in 11 cases, or 7 percent.   

In three cases of the bronchopneumonia type, smears from the pus droplets expressed from the small consolidated areas in the lungs showed innumerable small Gram-negative bacilli, and B. influenzae was cultivated, twice in pure culture and once with M. catarrhalis. Cultures from the heart's blood were negative in all three cases.   

Pneumococci from all cultures were typed 38 times.   

For cultures of B. influenzae during the first part of the epidemic, the surface of ordinary blood-agar plates was inoculated. Later the blood-agar was heated to 70° C. for 15 minutes, poured in plates, and the surface inoculated. The latter method proved much more effectual for the isolation of B. influenzae.   

Cultures from the lungs in 20 lobar pneumonia cases showed influenza and streptococcus viridans in 1, influenza and pneumococci in 2, pneumococci alone in 9, pneumococci and Streptococcus hemolyticus in 1, Streptococcus vividans alone in 1, and staphylococcus alone in 1. Cultures from the heart's blood in post-mortem cultures in 20 cases of lobar pneumonia showed influenza and Streptococcus viridans 1, influenza and pneumnococcus 2, pneumococci alone 7, pneumococci and Streptococcus hemolyticus 1, Streptococcus hemolyticus alone 1, Streptococcus viridans alone 1, staphylococcus alone 4. Thirty-eight cultures from pneumonia cases were typed with the result as follows: Pneumococcus Type II, 3; Type 11A, 1; Group IV, 34.   

The results of the bacteriologic examinations were disappointing and their value small in proportion to the amount of work entailed.   

B. influenzae
was a common organism during the epidemic, and was the one encountered most frequently in the group of selected severe respiratory infections. It was less frequently found than the pneumococcus in the series of 148 consecutive cases, and it was isolated only three times in the post-mortem cultures from the lungs and three times from the heart's blood in 20 lobar


pneumonia cases, twice with the pneumococcus and once with Streptococcus viridans. It was present, however, in cultures from the lungs in all of three fatal eases of bronchopneumonia, twice in pure culture and once with M. catarrhalis.   

Measles in this camp was not accompanied by an appreciable incidence of pneumonia and by practically no fatality. The influenza epidemic of the spring of 1918 was accompanied by considerable pneumonia, with a sharp rise in the mortality. The influenza epidemic in October, while accompanied by a distinctly higher fatality rate than in the spring, was not as serious as in many camps. The types of lesion appear to have been similar to those generally present with the exception that there was no evidence of the presence of intetstitial pneumonia with its high empyema incidence caused by the streptococcus. Streptococcus hemolyticus was found much less frequently than in most of the camps, pneumococcus Group IV being distinctly the predominant organism according to cultural findings. Cultural work was not sufficiently extensive to offer much of value on the subject of the epidemic etiology. The types of pneumonia found corresponded with the types found in other camps where the same or similar bacterial distribution occurred.   

An increase in respiratory-disease fatality preceded the epidemic wave of influenza in this camp by a considerable period. Whether this indicates an increasing virulence of organisms causing these infections it is impossible to determine. In many camps it preceded the epidemic wave by a very short time, usually a matter of days or weeks. Its wide separation in this camp suggests that it has no particular relationship to the increase in virulence of organisms causing the pandemic.


Data, other than the reports of sick and wounded, are not available to determine the types of serious infection of the respiratory tract prior to the influenza pandemic of the fall of 1918. These reports show a relatively low case fatality at the peak of incidence in January, 1918; 0.6 percent of the total respiratory disease incidence. The deaths were attributed to primary lobar pneumonia, and Group IV pneumonococci preponderated in the cultures from the cases. During February, 1918, the incidence of inflammatory respiratory diseases declined, but the case fatality rose sharply to nearly double that of January, or 1.10 percent. The occurrence of the streptococcus among the cases increased at this time, as did empyema.   

In April, there was a sharp increase in the case fatality rate to 3.73 percent, at which time streptococci were abundant in the cultures and empyemia was more frequent. Since the camp personnel had been practically constant for four months, this increase in the case fatality rate suggested an increase in the virulence of the infecting organisms. No available data reveal information about the influenza bacillus prior to the pandemic of the fall of 1918. The incidence and type of the acute repiratory infection suggest an influenzal wave both in the fall of 1917 and in April of 1918, particularly in the latter month.

y Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1817, 1918, and 1919.

82 CHART IX.- The incidence and fatality of the acute respiratory diseases at Camp Grant


The following discussion concerns the pandemic of the fall of 1918.z   

An epidemic of unusual virulence swept with great rapidity through several organizations in Camp Grant between September 21, and October 18, 1918 (approximately). During this time 9,037 patients were admitted to the base hospital, representing about one-fourth the strength of the camp, and of them 26 percent developed pneumonia. About 11 percent of the total admissions, or 43 percent of the total pneumonia patients died.   

From the throat cultures of the early admissions pneumococci were recovered with remarkable constancy; influenza bacilli were found exceptionally. Post-mortem examination of patients dying during the epidemic demonstrated regularly a massive bronchopneumonia. In cultures of the diseased lungs, heart blood, exudates, fluids, and other diseased foci of the body, pneumococci were the predominating organisms. Blood cultures of 90 patients in the hospital with pneumonia of the epidemic disease were positive in 50 percent, pneumococci being the only organisms recovered. Pneumococci in pure cultures were recovered from infected sinuses of the head, the middle ear, conjunctival and empyemic exudates. Inoculation of animals with pneumococci isolated during the epidemic demonstrated a high virulence of these organisms. Control throat cultures made of prisoners in the camp escaping the epidemic disease contained practically no pathogenic organisms.   

Of 21 positive blood cultures during life, 14 were Type II, 1 Type I, and 6 Group IV. Of 16 post-mortem, positive blood cultures 10 were Type II, and 6 Group IV. Of 13 cultures from the lung, post-mortem, 8 were Type II and 5 Group IV.   

During the epidemic 198 post-mortem examinations were made.   

A decided cyanosis of the face, neck, and extremities was observed frequently, and commonly there was a distinct yellow tinge to the skin and conjunctivae. Dryness of the skeletal muscles was usual. The parietal pleura commonly was hyperemic and glistening, or slightly dull, with loss of the normal smoothness, and covered by a scanty layer of fibrin. Turbid brown fluid, from a few to 300 c.c., was contained in each pleural cavity when the inflammation had penetrated into these serous spaces. Changes in the pericardial sac consisted largely in an increase in the amount of yellow fluid normally contained. A true serofibrinous or suppurative pericarditis was not seen until late in the epidemic, and acute changes in the heart valves were not observed in any post-mortem examination. The myocardium, and all the parenchymatous organs as well, presented a moderate or severe grade of cloudy swelling, with commonly disseminated areas of acute fatty changes. Subserous, petechial hemorrhages were common. The parabronchial lymph glands constantly were enlarged, very soft, sometimes with small areas of necrosis. The lining of the trachea and of the main bronchi was intensely hyperemic, and on the mucous membranes there was a thin milky gray, brown, or sanguineous fluid, sometimes frothy. In the deeper

zThe following statements of fact are based, in the main, on studies by Edwin F. Hirsch and Marion McKinney: Epidemic of Broncho-pneumonia at Camp Grant, III. Journal American Medical Asociation, Cbicago, 1918, lxxi, 1735. An Epidemic of Pneumococcus Broncho-pneumonia. Journal of Infectious Diseases, Chicago, 1919, xxiv, 594, and xxv, 394, respectively.


portions of the respiratory tree the bronchioles were lined by an intensely red mucous membrane covered by a relatively thin, gray, exudate. The consolidation of the lungs was extensive. When removed from the chest the lungs appeared voluminous, dark red or cyanotic, with no, or very little, exudate on the pleural surface, these membranes commonly being quite smooth and glistening. Underneath the pleura there were hemorrhages into the lung tissue from one to several centimeters in extent, as a rule, in addition to nodular consolidations which sometimes were so extensive as to form large confluent areas of firm tissue. Such extensive consolidation usually occurred in patients who had survived several days, those in whom the lungs were extremely hemorrhagic dying relatively early in the disease. Several of the lobes were involved. From the surfaces of the lung tissue large quantities of blood and bloody fluid escaped, and there were firm areas of red tissue slightly granular and moist, with edematous and hemorrhagic lung tissue between. As the epidemic progressed gray nodular consolidations often confluent were noted. On the cut surface the tissue was distinctly granular and moist, with a brown or brownish-red exudate.   

The yellow of the adrenal cortices was moderately diminished, or entirely absent, and in one case there were several small subcapsular hemorrhages and in each organ. As a rule, the spleen was increased in size from one-half to twice the normal, and the substance of the organ was soft and dark red. The biliary and mesenteric lymph nodes constantly were increased in size, the former much more so relatively than the latter, the substance moderately firm and red. Changes in the common bile duct, the portal vein, the splenic, and the upper portion of the superior mesenteric veins were not seen.   

The liver was enlarged, its capsule tense, and the tissue beneath mottled with disseminated areas of acute fatty changes and passive hyperemia. Focal necrosis was not recognized grossly. In the lining of the stomach there were multiple petechial hemorrhages.

In a limited number of cases in which the brain was examined there was found a moderate hyperemia of the pia-arachnoid, with dryness of the cerebral cortex. The spinal fluid of such patients usually without meningeal manifestations, commonly was turbid. In a limited number of examinations collection of a mucopurulent exudate was found in the sinuses of the face and middle ears.   

As the epidemic progressed, complications commonly following pneumonia were found post mortem, including serofibrinous pleuritis, empyema of the chest, acute serofibrinous pericarditis, acute suppurative pericarditis, and peritonitis. The character of the lung changes varied, too, in the later stages, and it was in the last few days of the epidemic that gray consolidations of the lungs were noted, and in a few of these extensive softening of the lung tissue. In some, lung changes grossly corresponded with descriptions of acute suppurative interstitial pneumonia. Mention should be made of cases in which the lung changes were comparatively insignificant, but in which disease of some of the bony spaces of the head occurred and subsequently or coincidently infection of the leptomeninges. In these the greatest interest centers in the bacteriologic examinations which are discussed later.


Following the decline of the epidemic, deaths occurred in the hospital fairly often with diseases primarily pulmonary, the lung changes of which differed from those observed during the epidemic. Usually such deaths followed an illness in the hospital of more than one week. In the lungs of those dying within the first or second week there were red or reddish-brown nodular peribronchial consolidations irregularly scattered throughout one or more lobes. The bron- chioles contained a white viscid exudate, the lining membrane being hyperemic and swollen. As this process became moderately advanced, the lungs contained many soft, yellow, peribronchial areas of necrotic tissue from 1-3 mm. in diameter, contrasting sharply with the red, air-containing or compressed lung tissue. Removal of the exudate from these areas exposed shallow crater-like peribronchial excavations in the tissue, without a definite limiting membrane, being covered by a yellow shaggy exudate. The tissue beneath the exudate was hyperemic or hemorrhagic. Lungs in which this peribronchial destruction of tissue was general were riddled by many abscesses of relatively small dimension without a definite limiting membrane. Such lobes sometimes were just masses of necrotic tissue held together by the supporting framework. Other lungs not as extensively diseased contained several or many scattered peribronchial abscesses of relatively small dimension having a thin pyogenic membrane.   

Clinically, the patients with extensive destruction of the lung tissue expectorated great quantities of yellow exudate and liquefied necrotic tissue, a phenomenon regarded by some as the emptying of an empyema through an eroded bronchus. The post-mortem examination demonstrated clearly that the yellow expectoration came from the lung itself and not from the pleural cavity.   

This acute suppurative interstitial pneumonia, however, was not the pneumonia that characterized the epidemic proper.
Microscopic examination of tissue from various places in diseased lungs disclosed irregularly distributed and developed inflammatory changes. In the alveolar spaces over small or wide areas corresponding with the gross changes, were red corpuscles in large numbers, and leucocytes in the ratio usually in the circulating blood. Such areas were distributed irregularly. Other alevolar spaces either in areas of hemorrhage or elsewhere contained endothelial cells in addition to the leucocytes, the number of endothelial cells increasing apparently with the progress of the lung lesion. Delicate strands of fibrin were present in such alveolar spaces, and also polymorphonuclear leucocytes, which gradually approached and exceeded the endothelial cells in number. In later stages the alveolar spaces contained numerous polynuclear leucocytes, slightly less numerous endothelial cells, fewer red corpuscles, strands and masses of fibrin, cellular detritus, and amorphous particles of coagulated material.   

In large confluent areas of consolidation the alveolar spaces regularly contained polynuclear leucocytes, endothelial cells, red corpuscles, and granular protein material as mentioned.   
The bronchioles invariably contained many polynuclear leucocytes, endothelial cells, red corpuscles, fibrin, and other products of an acute inflammatory reaction.   

The post influenzal respiratory diseases also were investigated. Starting in December, 1918, and continuing into March, 1919, 455 throat cultures of


patients admitted for influenza were studied on plain blood agar according to the methods used in the earlier studies. The results of studv of throat cultures of influenza patients showed Gram-positive, lancet-shaped diplococci, sometimes in chains, growing in green colonies on plain blood agar in 71.87 percent; hemolytic streptococci in 21.98 percent; nonhemolytic streptococci in 9.01 per cent; B. influenzae in 19.58 percent; and Micrococcus catarrhalis in 72.31 percent.
Of the Gram-positive, lancet-shaped diplococci, 169 pure cultures were isolated and studied more carefully; 19 were soluble in bile and fermented inulin, 43 were soluble in bile only, 45 fermented inulin only, and 32 gave no reaction with either of these tests.   

The anatomic changes observed after death, when the epidemic had subsided, were chiefly such as complicate or follow pneumonia. Empyema was the commonest, usually in patients in whom the purulent exudate in the chest had been drained. Acute suppurative pneumonia with multiple peribronchial abscesses without definite limiting membranes was fairly common. The amount of destruction in such lungs varied widely, in some practically all the respiratory tissue of a lobe or a lung having been destroyed. Hemolytic streptococci were recovered from the exudate of such lungs with great regularity. It is possible that all of these acute suppurative pneumonias were essentially such from the beginning, but there is reason for believing that some of them were superimposed on a primary pneumonia without suppuration, because the examination of sections from earlier stages of such lungs showed necrosis in the alveolar walls or ductuli alveolares, this necrosis being but a small part of a much larger consolidation with changes like those in pneumococcus infection. In addition acute suppurative leptomeningitis and acute thrombo-ulcerative endocarditis were observed a few times.   

Much different were the results in deaths from respiratory infections that in no way could be considered a complication of the epidemic disease. This group contains the largest number. Here the lungs were mottled by many, comparatively small, dark red areas, the increased firmness being due largely to the escape of blood into the substance. Large amounts of bloody fluid escaped from the cut surfaces, and the pleural cavities contained from 100 to 500 c.c. of thin reddish-brown fluid. Hemolytic streptococci were recovered in pure culture from the lungs and heart blood. In other cases the lungs contained irregular, nodular consolidations several centimeters in dimension, representing an older stage, and culturally yielding a similar bacterial flora. These nodules did not resemble in close detail the consolidations observed during the height of the epidemic in that the surfaces of the latter at the same stage of development were more coarsely granular without the yellow focal areas of necrosis, and the fluid expressed from the tissues was more viscid and grayish-brown rather than the reddish-brown of hemolytic streptococcus infection. The early acute hemolytic streptococcus pneumonia is also quite distinctive from the pneumonia of the epidemic.   

During the five months in which these observations were made, there were only three cases in which the typical lobar pneumonia was approached. All of these were noted after March 15, 1919. Consolidations of entire or


nearly entire lobes were observed, but these were confluent bronchopneumonia rather than lobar pneumonia. That lobar pneumonia should be submerged during the epidemic and for months after is of great interest. There seems to have been some fundamental change in the reaction of the host toward the infecting agent, some difference in the activity of the invading organism, or change in the interaction of both.   

Cultures of tissues and fluids taken post mortem afford valuable information in acute infectious diseases of the variety of pathogenic bacteria present, and correlate the lesions with the organism concerned. The methods used during the epidemic were continued, with slight modification, in this study. As a rule, the cultures were obtained within six hours after death, and usually were controlled by direct smears made at the same time, stained later by Gram's method and dilute aqueous fuchsin.   

During the period when complications of pneumonia were observed post mortem, and again later in a number of acute respiratory infections, hemolytic streptococci were recovered in mixed and pure cultures. With the ascendence of hemolytic streptococci in cultures, pneumococci became less frequent, although there were a few examinations during the first three months in which the lung changes and results of cultures corresponded well with those of the epidemic. In all the cultures from various places, influenza bacilli were found only occasionally with other organisms, never in pure culture. Pure cultures of Staphylococcus albus were recovered from diseased lung tissues, body fluids, and the spleen in two examinations.   

The lung culture results are briefly as follows for 49 of the 89 cases studied in which death probably had resulted from an acute respiratory disease after the epidemic or from lung complications of the epidemic disease: Hemolytic streptococci in 30; pneumococci in 20; Staphylococcus albus in 11; B. influenzae in 8; nonhemolytic streptococci in 2.   

These results may be misleading in that each of the various organisms listed as being present were not regarded in every instance as being the predominant organism. Of the remaining 40 examinations, 31 presented no immediate reason for culture of the lung, 3 others were not tabulated because of an active tuberculosis, while in the remaining 7 an acute suppurative leptomeningitis was demonstrated, the spinal fluid of 4 containing pneumococci, and of 3 hemolytic streptococci.   

Heart-blood cultures of the 89 necropsies gave the following results: Hemolytic streptococcus, pure, 26; pneumococcus, pure, 18; Staphylococcus albus, pure, 2; negative, 16; not cultured, 27.   
These results indicate that, with the decline of the epidemic and subsequently, hemolytic streptococci became important as invaders of tissues already diseased and frequently provoked disease in tissues not the site of a preceding change.
The post-mortem bacteriologic studies demonstrated the presence of hemolytic streptococci, pneumococci and staphylococei in diseased fluids and tissues.   

The fixed types of pneumococci, particularly Type II, were extremely important in this camp. Group IV apparently had a relatively high incidence during the early months, with an increase in streptococcus during the epidemic or increased incidence of acute respiratory disease in the month of April, in a


personnel which had been practically constant since September, 1917. Pneumococcus Type II was an extremely important infecting agent and evidently of very high virulence during the influenza epidemic. The incidence of streptococcus increased during the influenza epidemic. It continued important throughout the postepidemic period. While the influenza bacillus was found in some numbers, it is doubtful if the total incidence was revealed. Descriptions of many of the cases are quite typical of the changes in the lung produced by this organism when not accompanied by others. There was evidence, however, of other organisms in such lungs in each necropsy. Pneumococcus Type II may be extremely virulent and occasionally produces a completely hemorrhagic lung with rapid death without appreciable consolidation other than inflammatory edema; in fact death, with septicemia without more than slight localization in the lungs, has been reported. At the Army Medical Museum there are on file several protocols of Type II infection which were essentially cases of septicemia, with little evidence of localization other than an acute bronchitis of hemorrhagic type.   

It is noteworthy that, in spite of the marked increase of camp personnel during August and September, 1918, there was no corresponding rise in the measles rate. Measles practically disappeared in several camps during the influenza epidemic period, but rarely did this happen when, coincidently with the onset of the epidemic, considerable numbers of unseasoned troops were received. The possibility that the virus of influenza can act unfavorably on that of measles is suggested by such an occurrence as is portrayed in the history of Camp Grant.   

There are 73 necropsy protocols of cases dying of acute respiratory disease at this camp in the collections of the Army Medical Museum, 37 of which are accompanied by gross and microscopical specimens of lungs and other organs.   

Organisms morphologically and tinctorially identical with B. influenzae are found along the smaller air passages in well fixed material of cases dying within the first 7 to 10 days. Other organisms, Gram-positive cocci, are present, however, in all cases. There are several cases from which pneumococcus Type II was isolated, but these show definite pneumonic lesions bronchopneumonic in character, with considerable diffuse hemorrhage, a true hemorrhagic inflammation.

CAMP GREENE, N. C. aa   

The acute respiratory diseases following mobilization at this camp, aside from the high incidence of measles, were due to influenza and common respiratory diseases. Five hundred and sixty-five cases were diagnosed influenza in the month of December, 1917. The deaths for December were 24 from complications of measles and 20 from primary pneumonias, the latter undoubtedly representing the terminations of unfavorable character in the influenza outbreak. Another outbreak of respiratory diseases, during a time when measles was dropping in incidence rate, occurred with a peak in April, the majority of the deaths being charged to primary lobar pneumonia. Here, as in other camps, patients were admitted for pneumonic complications, the preliminary respir-

aa Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1917,1918, and 1919.


atory attack having occurred before their entrance into hospital and therefore not appearing as an admission diagnosis.   

Cases of pneumonia began to appear as soon as the base hospital was opened in September, 1917, but were comparatively few until about the middle of December.bb With the onset of the cold weather, however, they increased until, by the end of April, 1918, when the cases ceased to come in, there had been admitted 427 cases of lobar pneumonia.   

Of the 427 cases, there were 34 of lobar pneumonia which appeared as a complication of measles. The other 393 were primarily cases of lobar pneumonia sent in from their respective commands or from the regimental infirmaries.   

The cases as they appeared from September to the end of January did not differ from those one sees in young people in municipal hospitals. During December the cases began to show a tendency to involve more than one lobe. During February they seemed to assume a somewhat different character and were more severe. Besides the multiple lobe involvement, the patients were usually pale, showed no herpes labialis, had an increased pulse rate out of pro- portion to the ordinary cases, and in general they seemed to be in a more or less typhoid state. In practically all cases there was obtained a history of cold in the head or chest or exposure to cold from a few days to a week before the onset of the disease.   

More than half the cases gave a history of onset with a chill and pain in the side. Localization of the pain is important and in a good many cases the area of pain was watched for signs of consolidation which appeared a few days later.   

A good many cases did not begin with a chill, but had pain in the side with cough and expectoration. Of these cases there were first the very mild ones, and, second, some of those appearing during February with more of a typhoid state.   

In a considerable number of cases the onset began with symptoms of meningeal irritation. The patients complained of headache, were stuporous, had stiff neck and a Kernig sign. These cases were usually sent in with the diagnosis of meningitis and the necessity of confirming or disproving this promptly led to the tapping of the spinal canal. The spinal fluid obtained, however, was clear, showed no cell increase, but was under plus pressure. Usually the meningeal symptoms improved with the puncture, but at times they persisted for more than 24 hours. None of the patients showed at this stage any signs in the chest, but in a few it was possible, after making them cough repeatedly, to obtain rusty sputum, making spinal puncture unnecessary.   

Quite a number were first admitted to the surgical service with a diagnosis of appendicitis.   
The dullness in pneumonia was usually moderate. Rarely was there complete flatness and the dullness usually stopped half an inch or so short of the vertebral column, never reaching the median line; while the cases of fluid were

bb The following statements of fact are based, in the main, on: Pneumonia at Camp Greene: A Few Considerations from a Clinical Standpoint, by Herman Elwyn, Southern Medical Journal, Birmingham, 1918, xi, No. 12, 780-785.


nearly always flat on percussion over the vertebral column and the dullness usually crossed to the other side.   

There were all kinds of variations, in the course of the disease, from the very mildest cases, lasting not more than three days, to the most severe.
As soon as pneumonia cases began to appear in the hospital, pneumococcus type determination was requested from the laboratory. This was ascertained by injecting the washed sputum into the peritoneal cavity of a mouse. The mouse was killed in 8 to 12 hours, and an emulsion of the organisms, obtained from the peritoneal fluid of the mouse, was tested for agglutination with standard sera of the various types.   

Of the complications the most important was empyema. Besides the empyemas complicating lobar pneumonia, a great number of empyemas complicated the purulent bronchitis following measles, and there were some empyemas secondary to other infections. Up to March 1 the following were observed: (1) Empyemas following bronchopneumonia and lobar pneumonia after measles; total cases, 14; deaths, 10; mortality, 71.4 percent. (2) Empyema after measles without pneumonia; total cases, 29; deaths, 18; mortality, 62 percent. (3) Empyemas without measles or pneumonia; total cases, 14; deaths, 4; mortality, 28.5 percent. (4) Among the 393 cases of primary lobar pneumonia there were 45 cases of empyema, a percentage of 11.4. Of these cases 21 died; that is, 46.6 percent of the 45 cases.   

As there was a good deal of similiarity among these various empyemas, it is well to consider them together. Clinically they presented themselves in three varieties.   

In the first class were the empyemas following early cases of pneumonia. They were usually of a mild degree. They did not show very marked prostration. The pus was thick and creamy and showed in the aspirated diagnostic sample either no organisms or mixed pneumococci and streptococci.   

In the second class were the cases of empyema in which the clinical picture was not so mild. Among these were cases which developed during or following the course of pneumonia or measles or were admitted as such. They were of various grades of severity. They presented on aspiration a brownish or milky-white, purulent, rather thin fluid, containing in a thin smear of the noncentrifuged specimen a moderate number of short-chain streptococci, which the laboratory classified as either hemolytic or viridans. Often the fluid was so clear that it was repeatedly aspirated before the case was sent to the surgical service. But in every case where clear fluid showed the presence of organisms, and where it was hoped that operation could be avoided, the fluid eventually became purulent and drainage had to be resorted to.   

In the third class are included the very severe cases. These often appeared after the pneumonia had subsided; they appeared also in the course of measles, bronchitis, and in a few cases secondary to a tonsillitis, or without any apparent cause. They presented a clinical picture totally unlike any other cases of empyema.   

The onset was with a sudden sharp pain in the side or in the abdomen. The most striking cases were those after measles in which normal convalescence had seemed established and the patients were up and about the wards with


normal temperature. The face assumed a grayish color,was pinched and anxious, the pupils were moderately dilated. The temperature rose promptly to 103 or 104F. The pulse was rapid, usually about 140 ° to 160 °. Breathing was shallow and rapid and the breath sounds on the affected side were diminished or absent. The prostration was extreme. Within four to six hours after the onset, fluid could be obtained on aspiration over the base of the lung. The fluid increased rapidly and in 12 to 18 hours could be percussed above the angle of the scapula. Dyspnea and prostration were present; pulse rate increased to a marked degree; the pupils became completely dilated; the patient sweated profusely; and death often resulted within 24 to 48 hours.   

The fluid was thin, of a brownish color, cloudy and contained in a thin smear a great number of short-chain streptococci, and practically amounted to a pure concentrated culture of streptococci. The greatest mortality was among these cases, many of which were entirely too sick for a rib resection, so that repeated aspiration was resorted to. A few cases received antistreptococcus serum intravenously, but without benefit. Most cases were operated upon and quite a few recovered.   

From the data available it is difficult to determine the acutal type of pathology found in the lungs in the cases of the fall of 1917 and early months of 1918. Available data indicate that the pneumonias were rarely the characteristic fibrinous or croupous lobar pneumonia. Many lobes usually were involved, the picture varying in different parts of the same lung and in different lungs. Empyema was an extremely important complication and for the most part appeared to be due to one of the members of the streptococcus group. The last group of cases was quite typical of empyema following the breaking down of streptococcus lesions after the decline of acute inflammatory reaction in the lung, or of cases of interstitial pneumonia in which suppuration occurred late.   

Empyema was the most serious condition with which the personnel had to deal. A special problem occurred in cases in which the pus had become encapsulated, either before or after rib resection. Of these encapsulations the most difficult to detect were collections of pus between the heart and the lung, which were practically inaccessible to diagnostic puncture. In other cases the pus was localized over an upper lobe anteriorly or had burrowed underneath the great vessels.   

There were several cases of bilateral empyema. One patient, who was operated upon on both sides, recovered.   

Among the 427 lobar pneumonias there were 8 cases of pericarditis. Among these were three cases of dry pericarditis that recovered. The rest were hemorrhagic or purulent, all containing short-chain streptococci. There was one case with acute endocarditis.   

There were five cases of peritonitis. All died. The organism obtained from the peritoneal cavity upon autopsy was a streptococcus and not a pneumococcus. Clinically the cases were characterized by the absence or lack of prominence of abdominal pain. The symptoms which called attention to the peritoneal involvement were distention, stoppage of all bowel movement and of flatus, followed by persistent vomiting. In cases of distention not due to peritoneal inflammation, so frequently occuring in lobar pneumonia, there


was no vomiting. The autopsy in these cases revealed a thin, purulent fluid in the peritoneal and pleural cavities. Thin smears showed the presence of organisms.   

In three cases pneumococcic meningitis complicated lobar pneumonia. The meningitis appeared relatively late in the course of the pneumonia, and was characterized by irregular pupils, stiff neck, a Kernig sign, and a stuporous and restless condition. The pneumococcus was recovered from the spinal fluid. In one case antipneumococcus serum was injected into the spinal canal, but without any benefit. All three cases died.   

There were two cases of pneumonia occurring two months or more after recovery from empyema. Both came to autopsy, thus giving an opportunity to view the end result of a cured empyema. It was not at all encouraging. The pleura was markedly thickened, densely adherent, and it was astonishing to see to what an extent the lung had shrunk.   

Two cases were complicated during their convalescence by abscess of the abdominal wall. This was drained in each case and a pneumococcus was obtained from the pus.   

Two cases were followed by suppurative parotitis, which necessitated drainage.   

One case of pneumonia, which was complicated by one of the fulminating empyemas and which died within 24 hours after the onset of the empyema, showed several areas of gangrene in the lung.   

Of special interest were the cases of recurrent pneumonia. The following three are examples:   

Hospital case No. 4818, age 25, was admitted January 14, 1918, with lobar pneumonia of the left lower lobe. Temperature became normal on January 15 and remained normal until January 31. On the 1st of February he had a rise in temperature with consolidation of the right lower lobe, and on February 6 consolidation of the right upper lobe. Temperature became normal on February 15. February 18 he developed consolidation of the left upper lobe. This last consolidation was followed March 6 by signs of an abscess in the left upper lobe. Elastic tissue was found in the sputum. He died on March 14. No autopsy.   

Hospital case No. 6295 and 7406, age 20, had a chill on February 5, 1918, followed by fever and pain in the left side. Admitted February 7 with lobar pneumonia of the left upper lobe. Crisis on the 9th. Uneventful recovery. February 26, returned to duty in good condition. The same night he had a chill followed by fever and was readmitted to the hospital the next day, February 27, with signs of consolidation of the left lower lobe and part of the left upper lobe. Temperature became normal on March 4 and remained normal. Discharged to convalescent camp on March 24.   

Hospital case No. 6702, age 19, was admitted February 13 with acute bronchitis. February 17 he developed signs of consolidation of the left lower lobe. Temperature dropped to 100 ° F. on February 19 and continued irregularly between 99° and 100 ° until March 1, while the left lower lobe was gradually cleaning up. On March 2 he had a sudden rise of temperature to 104°, with signs of consolidation of the left upper lobe anteriorly. Temperature became normal on March 8, with the left upper lobe beginning to clear up. On March10 he had a rise of temperature to 105 ° F., and the next day signs of consolidation of the right upper lobe. Crisis occurred on the fourteenth day. Temperature was subnormal and in a few days became normal. Returned to duty.   

There were 34 cases of lobar pneumonia following measles. Of these, 17 died, that is, 50 percent. Contrast with these figures the mortality among the other cases of lobar pneumonia. Of the 394 cases, 68 died, that is, 17.2


percent. Of the 34 cases mentioned above, 11 had empyemna, 4 among which recovered. Here the problem was somewhat different. The course of events was as follows: The patient was admitted for measles and developed during the course the ordinary mild bronchitis accompanying measles. This gradually increased in severity, and after a time became purulent, involving every bronchus and bronchiole in both lungs. This was followed by consolidation of one or more lobes. Empyema complicated the pneumonia or occurred without the pneumonia.   

The difficulty in these cases was the purulent bronchitis. The patients coughed up pure pus for days and weeks. They became intensely dyspneic, cyanotic, and exhausted from the coughing. The absorption of toxic products from the bronchi must have been enormous and certainly must have contributed to the severity of the disease. When consolidation of a lobe supervened, this still increased the dyspnea and cyanosis and of course added to the severity. On the autopsy table the lungs showed besides the consolidation, which was of the lobar type, all bronchi and bronchioles filled with a thick yellow pus, which could be squeezed out under the slightest pressure.   

From this study it was learned that it is practically impossible to recognize clinically whether these cases of purulent bronchitis are accompanied by bronchopneumonia or not. Two cases of purulent bronchitis that died the same day had shown identical clinical pictures and had both been diagnosed bronchopneumonia. At autopsy one showed areas of bronchopneumonia in both lungs with the purulent bronchitis. The other showed the purulent bronchitis with not even the slightest trace of bronchopneumonia.   

During the influenza epidemic the mortality was due almost entirely to pneumonia and was so recorded by the medical personnel of this camp. One hundred and forty-one autopsies were done on cases dying during the influenza epidemic. Of these pneumonias 72 percent were diagnosed as of the lobar type.c   

The fact that clinically there seemed to be a sharp distinction between the simple cases without pneumonia and those resulting in pneumonia inclined the clinician to favor the view that the epidemic was one of a definite disease of which pneumonia was a distinct though prevalent complication.   

Clinically, the disease was one of sudden onset with symptoms seeming to show a severe initial general intoxication with little evidence of local source. The negro soldiers who were the first and most numerously attacked at this camp would frequently lie down wherever they happened to be when taken with the disease, having to be carried to the infirmary or hospital; headache and general pain in the lower lumbar or sacral regions were the almost general early complaints.   

Cough usually developed gradually on the second or third day. Coryza and lacrymation were infrequently noticed by the patient but in most cases objectively slight snuffling and slight conjunctival congestion were present. Expistaxis was frequent in early cases. A few patients had hematemesis or small intestinal hemorrhages. Many white patients on admission, showed

cc The following statements of fact are based, in the main, on: Influenza-pneumonia at Camp Greene, N. C., by Claude P. Brown and Francis W. Palfrey. New York Medical Journal, 1919, cx, 316-321; 368-372.


slight cyanosis of lips and ears. Somnolence was a conspicuous feature, almost all patients would apparently go to sleep as soon as they were put to bed, complaining of headache, if aroused, but otherwise lying perfectly quiet for long periods. No skin rashes were observed, although from two to three weeks after onset there was a general fine desquamation similar to that of measles, conspicuous especially on the skin of Negroes. The throat was often slightly red- dened but never to such a degree as to suggest the throat as an important seat of infection. Examination of heart and lungs in the stage of onset was uniformly negative. The spleen was not found enlarged, nor was there any abnormality of the abdomen, except in cases with vomiting in which there was often more or less vague abdominal tenderness, usually more marked in the upper portion. There was no general glandular enlargement which was not explicable by other existing causes. The urinary secretion was diminished as usual in acute febrile diseases, the urine concentrated and slightly albuminous. The white count was inconstant, as will be discussed under laboratory findings, together with blood cultures, sputa, and other bacteriological studies. The incubation period in one group of cases which could be traced to a probable source was five days.   

In clinical course the first day was marked by fever, general pain and prostration without other features. On the second and third days cough developed but at first without r les in the chest. On the third or fourth day in the milder cases the temperature was lower or normal and convalescence began. In a certain proportion of patients an apparent relapse occurred after one or two days, and in some patients several of these relapses occurred.   

The increased respiratory rate of beginning pneumonia was often of quite sudden appearance and in conspicuous contrast to the slow quiet breathing in uncomplicated cases. Thus while signs of consolidation were not infrequently found in patients with normal rate of respiration, all patients with rates approaching 30 a minute were examined with special care. Rusty or bloody viscid sputum was taken as evidence of pneumonia for purposes of isolation and treatment without waiting for signs of consolidation which regularly appeared later.   

The evolution of the symptoms and signs of pneumonia showed certain variations from those commonly observed in primary lobar pneumonia. Pleural pain, while present in certain patients, was relatively infrequent. Dyspnea while sometimes extreme was in many cases slight. The sputum, while typically rusty in many cases, in many others was light green and less tenacious. Whereas in primary lobar pneumonia it is exceptional to hear rales before bronchial breathing and then only a fine explosion of very fine crepitations at the end of inspiration, in these patients it was the rule to hear rather coarse crackling rȃles with somewhat diminished breath sounds for a day or more before bronchial breathing appeared. Dullness was often marked in the area of these rȃles while bronchial breathing was still absent. (Certain patients that died while this combination of coarse rȃles, dullness, and diminished breathing was present showed in the corresponding portion of the lungs dark red areas of consolidation of somewhat flabby jellylike consistency.) In practically all serious cases the signs of consolidation were extensive and in a great majority of the fatal cases this process was bilateral.


  Whereas in other hospitals bronchopneumonia predominated, in this hospital, with the exception of the variations noted above, the clinical course of the great majority of cases was typical of lobar pneumonia. With few exceptions the local physical signs of consolidation were distinct; few cases came to autopsy without local consolidation in part or in whole correctly predicted, and in the relatively few cases which proved to be bronchopneumonia the absence of clinical signs of consolidation had been noted as unusual. These cases of pneumonia tended to rather long course with successive appearance of new areas on consolidation. Their favorable termination was more often by rapid lysis than by sharp crisis.   

A tendency to delayed resolution was marked in a distinct proportion of cases, the physical signs and X-ray picture of consolidation persisting for one or more months. Exploratory punctures proved negative. The dullness and bronchial breathing would then gradually disappear but coarse crepitant r les would continue in the area for another long period after which complete return to normal was apparent.   

Serous pleural effusions of sufficient volume to be of clinical importance were occasionally found but were not numerous. Pericarditis was infrequent and pericardial effusions few. Empyema occurred in two classes, the early and the late. The first comprised turbid bacteria containing fluids appearing while acute pneumonic symptoms still dominated the clinical picture. Many of these cases were fatal, but it was believed that death, in these cases, was to be attributed to pneumonia rather than to empyema, the symptoms being pneumonic, and at autopsy the lungs being extensively involved in red or early gray hepatization. Experience with empyema in the previous year (1917) had led to the conviction that operation in the presence of an active pneumonia only aggravated an already severe condition and often accelerated a fatal outcome. When fluids even if becoming purulent and rich in bacteria were found, therefore, where it was believed that active pneumonia was still present, particularly if it involved the opposite side, it was the practice to depend upon aspiration only, deferring surgical drainage until a more favorable opportunity. The second class of cases comprised empyemas first developing after the pneumonia had run its course. In most of these the pus was encapsulated in rather small accumulations of relatively low toxicity. Some of these encapsulated areas gave one the impression of the fluid having accumulated within areas walled off by fibrous bands of adhesions which were due to an old pleurisy. Pneumothorax occurred in 12 patients. In eight which came to autopsy all but one had no tuberculosis or other lesion which could be found to account for the escape of air. Since acute emphysema with air vesicles of more or less prominence was not an uncommon finding in other cases it is probable that these collapsed lungs had previously contained such vesicles, the rupture of one of which had produced the pneumothorax. In all cases entrance of external air from aspiration was excluded. Wounding of the lung by exploratory puncture was not regarded as probable in any; in most no exploratory puncture had been made. In three of these patients pyopneumothorax developed. In two the pneumothorax was bilateral.


  A group of cases of considerable interest, studied by the radiologist, presented X-ray evidence apparently conclusive of lung abscess, tending to quite rapid spontaneous recovery by an absorption. From two of these, sterile nonodorous thin green pus was aspirated. None had profuse expectoration or foul breath. Local physical signs were slight or absent. All showed progressive disappearance of the lesion, became free of symptoms, and returned to duty. There were five cases of suppurative parotitis. In twelve cases of tuberculosis the disease may have been activated by influenza.   

As to the etiology of the epidemic, bacteriological data failed to establish the influenza bacillus of Pfeiffer as the infecting organism. The following studies of cases were made in the acute stage of influenzal symptoms, usually in the first 24 hours. In cultures from throat swabs, plated on defibrinated blood agar from 88 cases, Bacillus influenzae was present in 18, but never predominating and in most cases but few colonies, while the pneumococcus, streptococcus, and the usual repsiratory organisms were present in large numbers. In the cultures from these 88 cases the organisms observed and the number of cases were as follows: Pneumococcus, 20; hemolytic streptococcus, 19; nonhemolytic streptococucs, 28; streptococcus viridans, 45; micrococcus catarrhalis (group), 70; staphylococcus, 8; Bacillus influenzae, 8.   

Smears from the throat stained by Gram's method in 109 cases showed influenza-like organisms 63 times. The organisms observed (morphologically) occurred in the following number of cases: Bacillus influenzae, 63; pneumococcus 59; streptococcus, 35; Gram-negative bacillus (not influenza), 19; staphylococcus, 35; Vincent's bacillus and spirillum, 6; diptheroids, 10.   

Blood cultures were made in 101 cases; all proved negative. The blood counts taken within 24 hours after the first symptoms were no indication of impending pneumonia.   

As efforts to recover Bacillus influenzae resulted in few of them being found in pure culture or even predominating from sputum and throat swabs, further effort was made to recover them by animal inoculation. Sterile cotton swabs were passed up and back of the uvula, then with a downward stroke the swab was rubbed over the nasopharynx and pharynx, and was then withdrawn. This swab was shaken in about one c.c. of sterile bouillon which was then injected into mice, intraperitoneally. This was done in 12 cases; 6 of the 12 mice did not die; from the 6 mice that died cultures made from both the heart blood and peritoneum, all proved negative for influenza bacilli.   

There were 403 white blood counts taken, practically all from pneumonia patients during the period in which type determinations were made. These varied within such wide limits that they were of little significance except that those showing high blood counts were considered as showing great resistance. Two hundred and ninety-one blood cultures were taken; 19 of these were positive for pneumococcus.   

Following the great wave of the epidemic, sporadic cases of the disease continued to occur and patients with pneumonia continued to be admitted, many of whom undoubtedly were preceded by mild unrecognized influenza. There were 4,789 cases of influenza and 688 cases of pneumonia up to February 28, 1919, with 295 deaths.


In the period of the epidemic proper there were 4,595 cases of influenza, resulting in 626 cases of pneumonia. There occurred among these 45 cases of otitis media, specimens of which were sent to the laboratory. Pneumococcus was present in 20 cases.   

Deaths occurring September, 1918, to February 23, 1919, numbered 308, of which 141 came to necropsy. It is to be noted that 72.34 percent were due to lobar pneumonia. Lobar pneumonia, 102 cases; bronchopneumonia, 12 cases; miscellaneous (influenza was the primary cause for admission in 14 of these cases, and 13 deaths from other causes), 27 cases. The 167 cases on which necropsy was not performed were diagnosed almost entirely as lobar pneumonia, and in practically all in which necropsy was not performed it was dispensed with under stress of the epidemic because of the fact that the clinical signs were considered sufficiently clear to render it unnecessary.   
Of the 141 autopsies, in 102 cases the lungs showed the pathology of lobar pneumonia.   

At the Army Medical Museum the pathology of the respiratory disease at this camp is represented by 33 cases of pneumonia, 27 of which occurred during the influenza epidemic. Microscopical sections are available of the 27 cases and gross specimens of 11 of these. From a study of the reports quoted, the protocols, and the specimens, it is evident that in a large number of instances the cases called lobar pneumonia were not the typical fibrinous or croupous pneumonia, but irregularly spreading consolidations involving many lobes, with a predominance of the type more generally called confluent bronchopneumonia. Areas of typical hepatization, however, were unusually frequent, particularly during the influenza epidemic, and may possibly be accounted for by a relatively high incidence, as shown in the bacteriological examinations, of the fixed types of pneumococci. The specimens available for study show every type of pneumonic consolidation described during the epidemic; some of these are obviously lobular or confluent bronchopneumonias which were diagnosed as lobar pneumonia in the descriptive data accompanying these specimens.   

The bacteriology by culture is indicated in the material quoted. Gram-negative minute bacteria were present in sections of the well fixed material, particularly in cases predominantly of hemorrhagic type. Of the 27 cases in the influenza epidemic only 7 show in the descriptions accompanying them that a definite hepatization was present. In one of the seven cases, a typical consolidation of the lobar croupous type was present. In the remaining six, hepatized areas and bronchopneumonias of various types were described. In the influenza epidemic of the spring of 1918 streptococci were evidently an extremely important factor in the production of empyema and mortality, while during the influenza epidemic of the fall of 1918, these organisms were not predominent, empyema was rare (2 cases), and a large proportion of the fatalities appeared to be due to the fixed types of pneumococci, particularly Type II.   

The recording of cases of empyema without pneumonia is exceedingly interesting and suggests involvement of the pleura from infected mediastinal lymphatics, without extension of the process to the pulmonary parenchyma. The


sharp rise in respiratory diseases in the month of April, 1918, with a constant or decreasing personnel speaks for an increase of the virulence of the organisms not dependent upon an influx of new susceptible material.


The character of the respiratory diseases at this camp is of considerable interest not only because of their relatively low incidence but because of the very low case fatality of the group as a whole. The increase in respiratory diseases of the mobilization period showed a case fatality of 0.3 percent in November, 1917, dropping to 0.06 percent the following month. This is contrary to the experience in other camps, the majority of which showed an increase in the case fatality rate as the epidemic continued. An increase in case fatality in February, 1918, when it was 0.56 percent, preceded the increased incidence or epidemic peak in March, when, however, the case fatality dropped to 0.19 percent.   

Aside from this minor rise there was no spring epidemic, as seen in so many of the camps, but there was a sharp increase in case fatality to 1.67 in June, when 14 cases died from pneumonia, diagnosed as primary. This increase in case fatality appears to have occurred too long before the pandemic rise of September and October, 1918, to have any relation to it.   

No written data or necropsy protocols are available for study from this camp for the period preceding the pandemic.   

A thorough study of influenza and bronchopneumonia at this camp was made by the pneumonia unit.ee   
In making this study, the records of 407 cases of influenza, and the records and necropsy findings of 152 fatal cases of bronchopneumonia were critically reviewed. They were fortunate in having necropsy reports in practically every fatal case.   

The syptoms of influenza in the epidemic of the fall of 1918 were exactly similar to the symptoms of previous influenza epidemics. The onset, as a general rule, was attended with malaise, headache, chills or chilliness, diffused body pains, aching muscles and joints, and more or less severe prostration. The temperature and symptoms were out of all proportion to the physical findings; in fact, in the majority of cases (61.7 percent) the physical examination was entirely negative.   
During the epidemic at this camp, there were 7,088 cases of influenza, not including 1,126 cases showing definite signs of bronchopneumonia on admission. Of these 7,088 cases, 2,730 showed signs of bronchitis during the course of the disease, and bronchopneumonia developed in 858, an incidence of 38.3 percent for bronchitis and 12.1 per cent for bronchopneumonia. Considering the cases diagnosed on admission as bronchopneumonia, secondary to influenza, the incidence of bronchopneumonia was 24.2 percent.   
The course of the simple influenza was very short. In most instances the temperature, which on admission ranged from 100 º to 104 º F., returned to nor-

dd Source of information, except as otherwise indicated: Medical reports to the Surgeon General, 1917, 1918, and 1919.

ee The following statements of fact are based, in the main, on: Influenza and Broncho-pneumonia at Camp Lewis, by William J. Kerr, Hugh K. Berkley, and T. Homer Coffin. New York Medical Journal, 1919, ex, No. 4,133-140; 184-187.


CHART X.- The incidence and fatality of the acute respiratory diseases at Camp Lewis