Studies on the Relation of Pneumotropic Streptococci to Influenza Virus

On the basis of the bacteriologic studies made during the pandemic of influenza of 1918^2,11,12,35 and since,^5,17,34 and the many studies on influenza virus,^{1,3,6,1,9,10,31,32} there can no longer be any doubt that cultivable organisms, especially streptococci, and virus, as now understood, are etiologic in influenza and allied diseases. The almost universal presence of highly virulent streptococci or pneumococci and in certain epidemics the presence also of virus are almost certainly more than merely accidental or casual.

Both virus and specific types of streptococci have been demonstrated in cases of encephalitis^4,25,27 and poliomyelitis^23,28,29 and encephalitis^13,21,22 and poliomyelitis viruses^18,23,24 have been produced from neurotropic streptococci. Filtrable organisms have been obtained in studies of influenza.⁸ Spherical and oval bodies of varying size, in diplococcal and short-chain formation, have been demonstrated by me in uncentrifuged filtrates of experimental and “natural” encephalitis and poliomyelitis viruses. Both large and small forms have been demonstrated unstained by means of the electron microscope²⁰ and large forms have been demonstrated after special staining by means of the light microscope.^14,15,23 Similar appearing bodies have since been demonstrated by Sharp and his associates in the sediment of ultracentrifuged equine encephalomyelitis virus³⁰ (eastern and western) and in influenza virus.³³ The variation in shape, size and grouping of these bodies in encephalitis and influenza viruses resembled that of streptococci isolated by my methods from tissues containing encephalitis or poliomyelitis virus and from nasopharynx and blood of persons ill with acute epidemic influenza.

Experiments designed to determine the nature of the relationship of pneumotropic streptococci to influenza virus were undertaken first in 1937 and have been continued since, as opportunity was afforded. The method used and the results obtained are set forth in this report.

Methods

Cultures were made from nasopharyngeal swabbings and blood of persons ill with influenza; from milk supplies and from freshly fallen snow obtained during epidemics of influenza; from emulsions and filtrates of emulsions of lungs of mice and from allantoic fluid of embryonated chicken eggs that had been inoculated with streptococci, with “natural” influenza virus and· with emulsions of pneumonic lungs in serial passage experiments beginning with pneumotropic streptococci. These materials were inoculated into my highly favorable medium, dextrose-brain broth, and on blood agar plates. The destrose-brain broth, which affords a gradient of reduced oxygen tension, a low oxidation reduction potential and other properties highly favorable for the isolation, rapid growth and preservation of specificity of causative streptococci, consists of tall columns of 0.2 per cent dextrose broth (pH 7.2) to which pieces of fresh calf brain are added. One part of brain substance to about 6 parts of broth is used and the tubes or bottles are then autoclaved for twenty minutes at 17 pounds’ pressure (1.2 kg. per square inch). The primary growth in dextrose-brain broth was plated on blood agar to determine the type of streptococci present. All cultures were incubated at 34°C.

Serial dilution cultures¹⁹ in alternate tubes of dextrose-brain broth and dextrose-brain agar (0.2 per cent dextrose and 0.2 per cent agar) also were made of the suspected materials and of mixed primary cultures. Pure cultures of the streptococcus from different dilutions often were inoculated to compare virulence of streptococci that grew in low and high serial dilutions. Cultures from the end point of growth in these mediums were mostly used to inoculate animals and for other tests. The dilution of original inoculum of cultures used for inoculation of animals was not less than 10^-12 and usually was much higher after repeated serial dilution cultures in which the transferring wire was sterilized at every step.

Freshly isolated, pure cultures of streptococci, far removed from original source (10^-12 or higher), which had not been grown on aerobic blood agar plates, were inoculated soon after isolation into autoclaved chick-embryo medium. Mice were inoculated with cultures in this medium after seven to ten days’ growth and after storage for from ten months to two and a half years.

The chick-embryo medium was prepared from a mash (1 part), obtained by passing nineteen day hatching chick eggs, including the shell, through a meat chopper, and 7 parts of distilled water. The mixture was infused in the refrigerator for twenty-four hours and was stirred frequently meanwhile. It was then placed into tall columns in test tubes or bottles and autoclaved at 17 pounds’ pressure (1.2 kg. per square inch) for twenty minutes. While the medium was still hot, a layer of sterile liquid petrolatum was added. The chick-embryo medium, which does not turn acid, is highly favorable for rapid growth of streptococci. When cultures in this medium are kept at room temperature in the dark, streptococci remain viable, without transfer, for a long time.

The methods used by me for the demonstration of virus in filtrates of nasopharyngeal washings from patients having acute influenza were similar to those used generally by investigators in the virus field. Mice were inoculated intranasally, while they were under deep ether anesthesia, with from 0.04 to 0.06 mil. (cubic centimeter), depending on their size, of filtrates of nasopharyngeal washings and filtrates of chick-embryo medium cultures of streptococci. Similar amounts of young cultures of streptococci in dextrose-brain broth and of seven to ten day cultures in chick-embryo medium, undiluted and diluted 1:100 to 1:10,000, were inoculated. In the first passage two to six mice were inoculated with one or more cultures or filtrates of cultures of the streptococci and in subsequent passages four mice were inoculated routinely with emulsions or filtrates of emulsions of pneumonic lungs. Control mice were inoculated similarly with filtrates of nasopharyngeal washings of well persons, filtrates of autoclaved chick-embryo medium, emulsions and filtrates of emulsions of lungs of normal mice from the same stock as used for the test mice, and with sterile saline solution and sterile dextrose-brain broth.

Mice that succumbed were examined as soon after death as possible and those that survived were killed with ether in from three to ten days after inoculation. Berkefeld V, N and W and Seitz filters were used. Mouse-to-mouse passage was carried out with 5 to 10 per cent emulsions which had been made with mortar and pestle in a non-stacked bacteriologic hood. Pneumonic lungs of inoculated mice were removed in a sterile manner and emulsions were made in saline solution or dextrose-brain broth but most often in the supernatant portion of autoclaved chick-embryo medium. Only lungs from mice that had just died, or more often from mice that were killed with ether, three to ten days after inoculation were used. Emulsions of lungs for routine animal passage in series were made immediately or after the lungs had been kept in 50 per cent glycerin for twenty-four hours or longer. Emulsions sometimes were inoculated after they had been kept frozen for long intervals in solid carbon dioxide.

Cultures of the emulsions or of pieces of lung, of pieces of brain and spleen, and of the blood of mice were made routinely in freshly prepared (one to ten days) dextrose-brain broth or dextrose-brain agar (0.2 per cent agar) and on blood agar plates. Only healthy, active mice that had smooth fur and had been under observation for some time were inoculated. Inoculated mice were examined twice daily for symptoms. The cages in which inoculated mice were kept were disinfected between batches and stock mice were kept in clean, well-ventilated cages. Animals were not inoculated with “natural” influenza virus while serial passages of lung emulsions, beginning with cultures of streptococci, were made.

Results of Studies On Pneumotropic Virulence of Streptococci

A large number of mice were inoculated intranasally with cultures of streptococci (1) from nasopharynx or blood of persons acutely ill with influenza; (2) from emulsions or filtrates of emulsions of lungs of mice during serial intranasal passage of lung material, beginning with streptococci; (3) from emulsions or filtrates of emulsions of lungs of mice during serial intranasal passage, beginning with “natural” influenza virus, and (4), as controls, from sources remote from influenza or influenza virus. The mice that died during the day were examined immediately and those that died during the night were examined in the morning; mice that survived were killed with ether in from three to ten days and also were examined for lesions. The results of such inoculations are summarized in table 1. There was a high incidence of lesions of lungs and of isolations of streptococci from pneumonic lungs after inoculation of test material and a negligible incidence of these after inoculation of control material.

Table 1: Incidence of Lesions of Lungs and of Isolation of Streptococci from Lungs of Mice That Had Been Inoculated Intranasally with Streptococci from Various Sources

*These mice were inoculated during the course of many similar experiments and were anesthetized and examined twelve to twenty-one days after inoculation. 

Results of Successive Passage of Lung Material, Beginning with Streptococci Isolated in Studies of Epidemic Influenza

The incidence of lesions of lungs of mice and of isolations of streptococci from lungs on serial intranasal passage of emulsions through nine consecutive passages, and of filtrates of emulsions of lungs through seven consecutive passages, after primary intranasal inoculation of a streptococcus isolated from the nasopharynx of a person acutely ill with influenza during the epidemic of 1937, is shown in figure 1. Note (1) the high incidence of lesions of lungs after inoculation of the streptococcus in the first passage, and of emulsions of lungs of mice through five additional passages, and the diminution of lesions on further passage; (2) the higher incidence of lesions after inoculation of emulsions, as compared with filtrates, in the fourth, fifth and sixth serial passages; (3) the general increase in incidence of lesions of lungs on passage of filtrate and (4) the somewhat higher incidence, in most passages, of isolation of streptococci after inoculation of emulsions than after inoculation of filtrates. In this series of experiments, in which fifty-four mice were inoculated with cultures or emulsions of lungs, lesions of lungs developed in forty-six (85 per cent) and the streptococcus was isolated from emulsions of lungs in twenty-nine (54 per cent). Of seventy-nine mice inoculated with filtrate, lesions of lungs developed in fifty-five (70 per cent) and the streptococcus was isolated from emulsions of lungs in fifteen (19 per cent).

Fig. 1. Incidence of lesions of lungs and of isolation of streptococci from lungs of mice on serial intranasal passage of emulsions or filtrates of emulsions of pneumonic lungs after primary inoculation of a streptococcus isolated from the nasopharynx of a person ill with acute influenza (1937).

Results similar to those shown in figure 1 were obtained with each of thirteen other cultures isolated from twelve persons having acute influenza (six from nasopharynx and seven from blood) during 1937 to 1941, inclusive. A summary of these results is given in figure 2. Note the similarity of the curves depicting the incidence of lesions of lungs and isolations of streptococci after inoculation of emulsions and filtrates of emulsions of lungs, respectively, and the downward trend of the curve of isolations of streptococci on successive animal passages. The streptococcus or emulsion of lungs was inoculated intranasally into altogether 447 mice. Lesions of lungs developed in 277 (62 per cent) of these. Cultures were made from the lungs of 387 mice; the streptococcus was isolated from 167 (43 per cent). Fifty-two mice (not given in fig. 2) were inoculated with filtrates of eight dextrose-brain broth cultures and the lungs of forty-seven of these mice were cultured. The mice had been killed with ether and the lungs were examined and cultured in the usual manner. Lesions of lungs were absent and the streptococcus was not obtained in any of the cultures. The filtrates of emulsions of pneumonic lungs of mice that had received these cultures in the first passage were inoculated intranasally into altogether ninety-eight mice, in fifty-eight (59 per cent) of which pneumonitis developed, and successive filtrates were inoculated into large numbers of mice. Altogether, 420 mice received filtrates representing lungs of mice that had been inoculated in from the second to the tenth consecutive passage. Lesions of lungs developed in 227 (54 per cent). Cultures were made from the lungs of altogether 399 mice, in 101 (25 per cent) of which the streptococcus was isolated.

Fig. 2. Incidence of lesions of lungs and of isolation of streptococci from lungs of mice on serial intranasal passage of emulsions or filtrates of emulsions of lungs of mice after primary intranasal inoculation of streptococci isolated from nasopharynx or blood of twelve persons while they were having acute epidemic influenza (1937-1941). 

Results of Successive Passage of Lung Material, Beginning with Streptococci Isolated from Milk During An Epidemic of Influenza

In studies of milk supplies in relation to current epidemics, it was found that streptococci, true to type, were isolated commonly. Those streptococci isolated from milk, even from pasteurized milk, during epidemics of influenza had high pneumotropic virulence and characteristic curves of cataphoretic velocity resembling those of streptococci isolated from nasopharynges of persons ill with influenza. They were agglutinated specifically by the influenza anti-streptococcic serum.²⁵

The possibility that these strains might yield a filtrable infectious agent resembling influenza virus was studied. Altogether eight strains isolated from pasteurized milk, representing the supply of a small city, during an epidemic of respiratory infections resembling influenza, were inoculated intracerebrally into mice. The cultures were derived from unopened bottles of milk in four instances in homes at the time cases occurred and in four instances from bottles that were purchased in the open market. Serial dilution cultures were made alternately in dextrose-brain broth and dextrose-brain agar, at steps of 10^-4, of each specimen of milk. The transferring wire was sterilized before each transfer. A second serial dilution culture was made of the streptococci obtained at the end point of growth from each specimen. The streptococci from the end points of growth of the second serial dilution culture were used for intranasal inoculation of mice and represented a dilution not less than 10^-12 of original inoculum (milk).

The incidence of lesions of lungs and of isolations of streptococci from lungs of mice on successive intranasal passage of emulsions or filtrates of emulsions of pneumonic lungs, after primary inoculation of two strains of streptococci isolated from pasteurized milk supplies during an epidemic of influenza in 1944, is summarized in figure 3. Note the high incidence of lesions of lungs in both groups of mice and the downward trend of the curve of incidence of isolation of streptococci on successive animal passage. In the two groups of mice, consisting of fifty-four and fifty mice, respectively, which were inoculated intranasally with these two strains, lesions of lungs developed in thirty-three (61 per cent) and thirty-one (62 per cent), respectively, and streptococci were isolated from lungs of fifteen (28 per cent) and eight (16 per cent), respectively.

Fig. 3. Incidence of lesions of lungs and of isolation of streptococci from lungs of mice on successive intranasal passage of emulsions or filtrates of emulsions of pneumonic lungs of mice after primary intranasal inoculation of two strains of streptococci isolated from pasteurized milk supplies during an epidemic of influenza (1944).

A summary of results obtained with six additional strains of pneumotropic streptococci that were isolated from milk during an epidemic of respiratory infections in 1944 is shown in figure 4. Note (1) the persistently high incidence of lesions of lungs that developed during the course of fifteen successive animal passages; (2) the prompt drop from a high incidence of isolation of streptococci in the first passage, in which streptococci were inoculated intranasally, to a low incidence of their isolation on successive inoculation of emulsions or filtrates of emulsions of pneumonic lungs. Of the 319 mice inoculated, lesions of lungs developed in 216 (68 per cent). Of 201 lungs from which cultures were made, thirty-four (17 per cent) yielded streptococci.

Fig. 4. Incidence of lesions of lungs and of isolation of streptococci from lungs of mice on serial intranasal passage of emulsions or filtrates of emulsions of pneumonic lungs after primary intranasal inoculation of six pneumotropic strains of streptococci. Three strains were isolated from the nasopharynges of patients who were having influenza and three were isolated from pasteurized milk supplies during an epidemic of influenza (1944).

Isolation of Pneumotropic Streptococci from Snow During An Epidemic of Influenza

During studies on the isolation of specific types of streptococci and streptococcal precipitinogen from air in relation to current epidemics, the results of which will be published elsewhere, cultures were made from rain and snow. The strain of streptococcus isolated from one specimen of snow during the onset of a sharp epidemic of influenza had such high pneumotropic virulence that it was studied extensively. Rabbits and mice were inoculated with the streptococcus as isolated and after storage in cultures in chick-embryo medium which had been kept in the dark at room temperature. The stored cultures were studied during and remote from epidemics of respiratory infections that occurred during the course of two and a half years.

Four samplings of fresh snow were collected in sterile glass jars the morning after a snowfall of 4 inches (10 cm.) had occurred during the night of February 8, 1942. There had been no snow on the ground for some weeks. The snow was obtained from the roof of a sixteen-story building, from ground level in the city, from ground level 2 miles (3 kilometers) out in the country and from ground level in a park at the outskirts of the city. No person or animal had been near the places where samples were taken. Collections were made against the breeze to eliminate the possibility of contamination from the breath or clothing of the person collecting the samples. The edges of the jars and screw caps were not touched by the gloved hands of the person collecting the snow. At each of the four places six jars, each having a capacity of 120 mil., were loosely filled with snow and the caps were firmly replaced immediately. The snow was allowed to melt at room temperature in the laboratory and each jar yielded 20 c.c. of decidedly turbid water. The appearance of the water was similar to that of water collected during first rains after dry spells in summer.

From this water, cultures were made, in a non-stacked bacteriologic hood, into tubes and bottles of dextrose-brain broth and on blood agar plates, inoculating heavily around the brain substance at the bottom of the tubes or bottles. The dextrose-brain broth had been steamed and then cooled immediately before inoculation. Precipitation tests were made with the water, which had been brought to isotonicity, (used as antigen) and antiserums prepared in horses with streptococci isolated in studies of influenza, encephalitis, poliomyelitis, arthritis and ulcerative colitis, and with control normal horse serum.

The cultures on blood agar plates from each sampling of the melted snow yielded Bacillus subtilis only. The cultures in dextrose-brain broth from each sampling yielded streptococci in pure form or in mixture with Bacillus subtilis. Pure cultures of streptococci were obtained from the mixed cultures by making serial dilution cultures alternately in dextrose-brain broth and soft dextrose-brain agar.

One rabbit was inoculated intracerebrally and two mice were inoculated intranasally with pure cultures of streptococci in dextrose-brain broth from each of the four samplings. Hemorrhagic edema or bronchopneumonia developed in all of the rabbits and in seven of the eight mice.

Just as growth of a long-chained streptococcus was occurring at the bottom of one tube of dextrose-brain broth that had been inoculated with the melted snow from the roof of the sixteen-story building, 1.2 mil. of the culture was inoculated intraperitoneally into each of two mice, 6 drops of the culture were given intranasally to each of two mice while they were under deep ether anesthesia and two rabbits were inoculated intracerebrally with 0.1 mil. of a 1:10,000 dilution of the culture. The mice that received the intraperitoneal inoculation died in twenty-four hours. Turbid pleural fluid and mild peritonitis were present on immediate examination. The two mice that had received intranasal inoculation were killed with ether twenty-four hours after inoculation. Severe hemorrhagic edema of lungs was found in one; the lungs of the other mouse were normal. The two rabbits died from hemorrhagic edema of lungs and bronchopneumonia within forty-eight hours after inoculation. The streptococcus was isolated from the blood, brain and lungs of both rabbits and from the lung or pleural fluid of the mice in which lesions had developed.

Results of Successive Passage of Lung Material, Beginning with the Streptococcus Isolated from Snow

The strain of streptococcus which had been isolated from the snow obtained from the roof of the sixteen-story building was subjected to two series of successive serial dilution cultures. Both series of serial dilution cultures were made alternately in tubes of dextrose-brain broth and dextrose-brain agar, each tube containing approximately 15 mil. of medium. Approximately 1.5 c.mm, of adherent culture was transferred serially with a nichrome wire. The transferring wire was shaken and twirled vigorously in each successive tube. In the first series of dilution cultures the wire was sterilized before each dilution. A pure culture of the streptococcus grew in dextrose-brain broth in the third tube, which represented a dilution of 10^-12. No growth occurred at higher dilutions. This culture was then subjected to a second series of dilution cultures, alternately in dextrose-brain broth and dextrose-brain agar, in which the transferring wire was not sterilized or heated between transfers. Pure cultures of the streptococcus grew in each of fifteen dilutions but not at higher dilutions. Autoclaved chick-embryo medium layered with liquid petrolatum, contained in 4 ounce (120 mil.) bottles sealed with rubber stoppers, was inoculated with a pure culture of the streptococcus from the end point of growth in the second serial dilution culture. After incubation overnight at 34°C, the subculture in chick-embryo medium was stored in the dark at room temperature.

Two mice each was inoculated intranasally with streptococci that grew in the third, fifth, seventh, ninth, eleventh, thirteenth and fifteenth dilutions of the second serial dilution culture; two mice were inoculated with streptococci that grew in chick-embryo medium which had been inoculated with a single colony from dextrose-brain agar in the fourteenth dilution; six mice were inoculated with streptococci, in the third subculture, that grew in dextrose-brain broth from the first serial dilution at 10^-12 and that had been passed through a rabbit and two mice were inoculated with this same strain in the fourth subculture in dextrose-brain broth. Lesions of lungs developed in twenty-one of the twenty-four mice inoculated. The streptococcus was isolated from the lungs of fourteen mice. The incidence of lesions of lungs or pleurae and of isolations of streptococci from lungs of inoculated mice on successive passages through mice of emulsions or filtrates of emulsions of pneumonic lungs of these mice is summarized in figure 5. The infectious agent which was obtained was passed successfully through fifteen serial passages in mice until early summer when it was lost in the sixteenth serial passage. Lesions of lungs developed in ninety-three (67 per cent) of 138 mice inoculated and the streptococcus was isolated from the lungs of twenty-six (19 per cent).

Fig. 5. Incidence of lesions of lungs or of pleurae of mice and of isolation of streptococci from lungs following inoculation of streptococci isolated from freshly fallen snow during an epidemic of influenza and following inoculation of emulsions or filtrates of emulsions of pneumonic lungs of mice on successive serial passage (1942).

Direct intracerebral inoculation of rabbits and intranasal and intraperitoneal injection of the water from the snow and of filtrates proved negative but precipitation tests with the influenza anti-streptococcic serum gave positive results, while tests with control antiserums gave negative results.

In vitro changes in cataphoretic velocity of streptococci which I have observed to occur during epidemics have been reported.¹⁶ Changes in virulence and in serologic properties of streptococci, which occurred during epidemics, have been found to occur on storage of chick-embryo medium cultures. Because of the precise results obtained with the streptococci on isolation from snow, a study was made of the streptococcus from snow which had been stored in chick-embryo medium. Over a period of two and a half years, during and remote from epidemics of respiratory infections, subcultures from the stored cultures were made in dextrose-brain broth and the young cultures were used for intranasal inoculation of mice. Results of these studies and those obtained on primary passage of the streptococcus on isolation from snow are summarized in table 2. The incidence of pneumonitis and of isolations of streptococci from lungs of mice was uniformly high on isolation and subsequently during current epidemics of respiratory infections and was in proportion to the severity of these epidemics. During summer when respiratory infections were absent, the incidence of lesions of lungs and of isolation of streptococci from lungs of mice was negligible.

Table 2: Studies of Specificity of Streptococci on Isolation from Snow  and After Storage in Chick-Embryo Cultures in Relation to the Seasonal Presence or Absence of Respiratory Infections

During the sharp epidemic of respiratory infections resembling influenza which occurred in January and February, 1943, one year after storage of the chick-embryo culture of the streptococcus from snow, a second series of experiments was undertaken in an attempt to produce the filtrable infectious agent again from the stored culture of streptococci that was still found to be highly pneumotropic (table 2). The results obtained in the first passage after inoculation of a young subculture of the streptococcus that grew in dextrose-brain broth, and in seventeen subsequent passages of emulsions or filtrates of emulsions of pneumonic lungs, are summarized in figure 6. Emulsions of lungs were inoculated in ten passages and filtrates in seven passages, both of which produced a high incidence of lesions of lungs. Lesions of lungs were found in seventy-six (76 per cent) of 100 mice inoculated and the streptococcus was isolated from twenty-five (25 per cent). There was a fairly close correlation between incidence of lesions of lungs and isolation of streptococci from lungs during the first eleven passages and in the sixteenth passage. The streptococcus was not isolated in the ninth, eleventh, twelfth, thirteenth, fourteenth, fifteenth, seventeenth and eighteenth passages, despite the fact that lesions of lungs occurred in high incidence. In the last three passages, the mice were inoculated in March during a sharp recurrence of an epidemic of respiratory infections and lesions of lungs developed in all. Direct intranasal inoculation of the stored culture of the streptococci in chick-embryo medium and of filtrates sufficed to produce pneumonitis in a series of mice.

Fig. 6. Incidence of lesions of lungs or pleurae of mice and of isolation of streptococci from lungs following inoculation of streptococci, isolated from freshly fallen snow during an epidemic of influenza, after storage in chick embryo cultures for one year and on successive passage of emulsions or filtrates of emulsions of pneumonic lungs of inoculated mice (1943).

The incidence of lesions of lungs and of isolation of streptococci from lungs of mice on successive intranasal passage of six strains of “natural” adapted influenza virus kindly supplied by others, and the incidence of lesions of lungs and of isolation of streptococci from lungs of mice on successive intranasal passage of emulsions of pneumonic lungs, after primary intranasal inoculation of streptococci isolated from five strains of “natural” influenza virus, are summarized in figure 7. Note (1) the somewhat higher incidence of lesions of lungs and a slightly lower incidence of isolation of streptococci from lungs following inoculation of emulsions of lungs, beginning with adapted virus, and (2) the striking similarity between the curves indicating the incidence of lesions of lungs and of isolation of streptococci on successive passage of emulsions of lungs, beginning with “natural” virus and those beginning with the streptococcus isolated from “natural” virus (fig. 7). The results summarized in this figure resemble closely those obtained on successive passage of emulsions of lungs or filtrates of lungs, beginning with streptococci isolated in studies of influenza (figs. 1 and 2), beginning with streptococci isolated from milk (figs. 3 and 4) and beginning with streptococci isolated from snow during an epidemic of influenza (figs. 5 and 6).

Fig. 7. Incidence of lesions of lungs and of isolation of streptococci from lungs of mice on successive intranasal passage of “natural” adapted influenza virus and of experimental virus derived from streptococci isolated from “natural” influenza virus.

Of 235 mice inoculated intranasally with emulsions of pneumonic lungs, beginning with “natural” virus, lesions of lungs developed in 164 (70 per cent) and fifty-eight (25 per cent) yielded the streptococcus in cultures from lungs. Of the 139 inoculated intranasally with the streptococcus isolated from “natural” influenza virus in the primary passage and with emulsions of lungs in subsequent passages lesions of lungs developed in seventy-six (55 per cent) and forty-one (29 per cent) yielded the streptococcus.

The results of intranasal inoculation of filtrate of nasopharyngeal washings from persons ill with acute influenza and the results of serial intranasal passage in mice of lung material, beginning with streptococci and beginning with “natural” influenza virus, obtained throughout these experiments according to the number of animal passages, in contrast to results following inoculation of control material, are summarized in table 3. Influenza virus was obtained from sixty-eight (26 per cent) of 259 mice that had received filtrates of nasopharyngeal washings from six of thirty patients having acute influenza during an explosive outbreak. Repeated attempts to obtain virus from cases of milder types of epidemic respiratory infection proved unsuccessful. A uniformly high incidence of lesions of lungs and a variable incidence of isolation of streptococci from lungs is shown in serial passage experiments beginning with streptococci and natural influenza virus. Thus pneumonitis developed in altogether 1,130 (59 per cent) of 1,900 mice and the streptococcus was isolated from pneumonic lungs in 461 (29 per cent) of 1,601 lungs cultured following inoculation of the streptococcus in the first passage and the streptococcus, emulsions or filtrates of emulsions of pneumonic lungs in subsequent passages. Pneumonitis developed in 164 (70 per cent) of 235 mice receiving “natural” influenza virus in the first passage and emulsions of pneumonic lungs in subsequent animal passages and the streptococcus was isolated from fifty-four (23 per cent). The incidence of lesions of lungs and average incidence of isolation of streptococci following inoculation of pneumotropic streptococci and following inoculation of lung material, beginning with streptococci and “natural” virus, respectively, were strikingly similar, whereas among 712 mice inoculated with control material the incidence of lesions of lungs and of isolation of streptococci from lungs was negligible.

Table 3: Summary of Results of Studies on the Relation of Pneumotropic Streptococci to Influenza Virus

After fixation in 10 per cent formalin of lungs and tracheas of a large number of animals that had been inoculated, sections were made and examined microscopically. The lesions of tracheas and of lungs of rabbits after intracerebral inoculation of pneumotropic streptococci that had been freshly isolated from blood and nasopharynges of patients during the acute stage of influenza, from snow during an epidemic of influenza and from “natural” influenza virus are well illustrated in figures 8, 9, 10 and 11. The lesions of lungs obtained in mice after intranasal inoculation of the filtrable infectious agent, after serial passage of pulmonary material beginning with pneumotropic streptococci and after intranasal inoculation of “natural” influenza B virus are illustrated in figures 12 and 13.

Fig. 8. Alveolar dilatation and distortion, interstitial congestion and hemorrhage and alveolar edema of the lung of a rabbit that had been inoculated intracerebrally with 0.1 mil. of a 1:1,000 dilution of a dextrose-brain broth culture of streptococci isolated from the blood of a patient having acute influenza (hematoxylin and eosin stain, X85).

Fig. 9. Congestion, edema, infiltration and hemorrhage of mucous membrane of the trachea of a rabbit that had died of hemorrhagic edema of lungs forty-eight hours after intracerebral inoculation of 0.1 mil. of a 1:10,000 dilution of a dextrose-brain broth culture of streptococci. The streptococci were isolated from freshly fallen snow during the onset of an epidemic of influenza. The culture used for inoculation had been obtained from a single colony of streptococci in dextrose-brain agar after four successive serial dilution cultures. This represented a dilution of the original inoculum of at least 10^-56 (hematoxylin and eosin stain, X 55).

Fig. 10. Congestion, hemorrhagic edema and infiltration of the lung of the rabbit referred to in figure 9 (hematoxylin and eosin stain, X 85).

Fig. 11. Alveolar distortion, hemorrhagic edema, bronchitis and alveolar and interstitial infiltration of the lung of a rabbit that had died forty-eight hours after intracerebral inoculation of 0.1 mil. of a 1:10,000 dilution of a dextrose-brain broth culture of streptococci isolated from influenza B virus (Lee E-3 strain). Note the resemblance in the lesions of the lungs of this rabbit to those shown in figure 10 (hematoxylin and eosin stain, X 85).

Fig. 12. Alveolar distortion, congestion and alveolar and interstitial infiltration of the lung (pneumonitis) of a mouse four days after intranasal inoculation of an emulsion of the pneumonic lung of a mouse, representing the fifth animal passage of pulmonary material beginning with a streptococcus isolated from the nasopharynx of a patient having acute influenza (hematoxylin and eosin stain, X 85).

Fig. 13. Alveolar distortion, congestion and alveolar and interstitial infiltration of the lung (pneumonitis) of a mouse four days after intranasal inoculation of influenza B virus (PR-8 strain) in the form of an emulsion of the pneumonic lung of a mouse (hematoxylin and eosin stain, X 85).

In order to determine whether the filtrable infectious agent obtained from pneumotropic streptococci resembling influenza virus was infective for embryonated chicken eggs, as is “natural” influenza virus, emulsions and filtrates of emulsions of lungs of mice, after successive passage of lung material beginning with streptococci, and after successive passage of “natural” influenza virus, were inoculated in parallel manner into the allantoic sac of ten to twelve day embryonated chicken eggs. The allantoic fluid of embryonated eggs in which the embryo had died, and, after from four to six days, of those in which the embryo had survived was, in turn, inoculated intranasally into mice and into the allantoic sac of embryonated eggs. The incidence of lesions of lungs and of isolations of streptococci from pneumonic lungs of mice, on egg-to-mouse and mouse-to-mouse passage, and the incidence of death of embryos and isolation of streptococci from allantoic fluid, on mouse-to-egg and egg-to-egg passage, are summarized in table 4. The incidence of lesions of lungs and that of isolation of streptococci on egg-to-mouse and mouse-to-mouse passage of the filtrable infectious agent (69, 87, 8 and 7 per cent, respectively) and of “natural” influenza virus(74, 81, 14 and 15 per cent, respectively) were similar but the filterable infectious agent proved more virulent than the “natural” influenza virus on mouse-to-egg and egg-to-egg passage. The mortality of embryos was higher after inoculation of the experimental filtrable infectious agent (49 and 72 per cent, respectively) than after inoculation of “natural” influenza virus (14 and 5 per cent, respectively). The incidence of isolation of streptococci from allantoic fluid of inoculated eggs was higher than from pneumonic lungs of inoculated mice.

Table 4: Incidence of Isolation of Streptococci from the Filtrable Infectious Agent and from “Natural” Adapted Influenza Virus After Passage from Egg to Mouse, Mouse to Mouse, Mouse to Egg and Egg to Egg, Respectively

Agglutination of Washed Erythrocytes of Chickens

A study was made, using the method of Hirst as modified by Rickard and his associates,¹⁰ of the agglutination of washed erythrocytes of chickens (1) by one strain each of influenza virus A and B in allantoic fluid of embryonated chicken eggs (kindly furnished by Dr. E. R. Rickard), (2) by centrifuged emulsions or filtrates of emulsions of pneumonic lungs of mice and (3) by filtrates of allantoic fluid of infected embryonated chicken eggs. The mice and eggs had been inoculated with pneumotropic streptococci, with emulsions or filtrates of emulsions of pneumonic lungs of mice or with infected allantoic fluid of eggs during serial passage of infective material, beginning with pneumotropic streptococci and influenza virus, respectively. Influenza virus A and B, as obtained and after passage through embryonated eggs, caused prompt agglutination of washed erythrocytes of chickens. Emulsions or filtrates of emulsions of lungs of mice that had received inoculations of pneumotropic streptococci or the infectious agent derived from streptococci on serial passage of lung material, caused only slight or no agglutination. The agglutinating action of influenza virus A and B, in emulsions or filtrates of emulsions of pneumonic lungs, was greatly reduced, or was lost, after several consecutive passages of the virus through mice.

Immunization of Mice Against Influenza Virus

Since the streptococci which I have isolated from adapted influenza virus and from the lungs of mice during successive passage of the infectious agent derived from streptococci resembled those which are associated with epidemic influenza, attempts were made to immunize mice against the streptococcus and the virus with vaccines prepared from streptococci freshly isolated in studies of influenza. Streptococci that had been freshly isolated in dextrose-brain broth and preserved in the glycerin saline solution menstruum were used to prepare the vaccines. The organisms in suspensions in saline solution, containing 20,000,000,000 streptococci per milliliter, were killed either by heating at 70°C for an hour, or by treating with 0.25 per cent solution of formalin at room temperature for twenty-four hours. The suspensions containing formalin then were washed twice in saline solution and resuspended in equivalent amounts of saline solution. To both heat-killed and formalin-treated suspensions, 0.2 per cent phenol was added as a preservative.

For intranasal immunization, 0.06 mil. of the respective vaccines, containing 20,000,000,000 organisms per milliliter, was instilled into the nostrils of mice while they were under deep ether anesthesia. This procedure was repeated on three successive days. The test material was inoculated intranasally in like doses five, ten or twelve days after the third daily instillation of vaccine. For intraperitoneal immunization the respective vaccines were diluted 1:10, and 0.1, 0.2 and 0.5 mil. were inoculated on three successive days. The test material was inoculated intranasally five to ten days after the third intraperitoneal injection of vaccine. Three, four or seven days after the test inoculations in the different experiments, the mice that survived were etherized and examined for lesions and cultures were made.

Immunized and control mice were inoculated intranasally in parallel manner with approximately 0.06 mil. of the supernatants or filtrates of 10 per cent emulsions of pneumonic lungs in saline solution, containing the infectious agent obtained from streptococci and “natural” influenza virus, respectively. The results in a series of experiments were so similar that the summary given in table 5 will suffice to illustrate. A high degree of protection was afforded by the two methods of immunization (intranasal and intraperitoneal) against both the filtrable infectious agent obtained from streptococci and “natural” influenza virus. The incidence of deaths, that of lesions of lungs and that of isolations of streptococci from lungs were consistently less in the immunized group than in the control group.

Table 5: Immunization of Mice with Influenza Streptococcic Vaccine Against Intranasal Inoculation of A Filtrable Infectious Agent Derived from Streptococci and of “Natural” Influenza Virus

Neutralization of Virus by Influenza Anti-streptococcic and Convalescent Influenza Serum

If the filtrable infectious agent obtained from streptococci, which resembled influenza virus and the “natural” virus are identical then they should be neutralizable in like manner with immune serums. Accordingly a series of experiments was carried out in which the neutralizing power of influenza anti-streptococcic and convalescent influenza serum were tested in parallel manner against streptococci isolated in studies of influenza, from the infectious agent derived from streptococci and from “natural” influenza virus. One part of suspension in saline solution, containing 20,000,000,000 streptococci per milliliter and 1 part of supernatant or filtrate of a 10 per cent emulsion of pneumonic lungs, containing the infectious agent or “natural” virus, were treated for one and a half hours at 35°C and in the refrigerator overnight with 9 parts of the respective serums. Two batches of influenza anti-streptococcic serum, several pools of serum which had been obtained from fifteen patients who were convalescing from influenza and normal horse serum were used. The serum-treated suspensions were then inoculated intranasally in parallel manner into a series of mice which were under ether anesthesia. Great care was taken that the depth of anesthesia and the amounts of inoculum in the test and control groups were comparable. The mice in the different experiments were killed with ether in from three to ten days after inoculation and they were examined immediately for lesions. Cultures from lungs, spleen and brain were made routinely in dextrose-brain broth and on blood agar plates.

A summary of these experiments is shown in table 6. In order to be able to compare readily the incidence of lesions of lungs and of isolations of streptococci in the test and control groups the results are given in percentages. Among the respective groups of mice that had received materials containing the filtrable infectious agent and “natural” influenza virus which were treated with the different serums, the average incidence of lesions of lungs and that of isolation of streptococci from lungs, spleen and brain were uniformly lower, often much lower, in the groups that received material treated with influenza anti-streptococcic serum and convalescent serum than in those that received the corresponding material after treatment with normal human and normal horse serum. Cultures from lungs, especially emulsions of lungs, occasionally yielded a few colonies of indifferent streptococci on blood agar plates but those from spleen and brain never yielded streptococci on blood agar. The statistical data, while striking, do not express adequately the difference in results in the different groups of mice. The lesions not only developed more often but were correspondingly larger in the test groups than in the control groups.

Table 6: The Neutralizing Action of Influenza Anti-streptococcic Serum and Influenza Convalescent Serum on Streptococci Isolated from Nasopharynx or Blood of Persons Having Influenza, On the Filtrable Infectious Agent Derived from Pneumotropic Streptococci and on “Natural” Influenza Virus

Summary and Comment

By the use of special methods, streptococci having high pneumotropic virulence were isolated from nasopharynges and blood of patients during the acute stage of influenza, from a milk supply and from freshly fallen snow during epidemics of influenza, and from pneumonic lungs of mice that had been inoculated with pneumotropic streptococci, with the filtrable infectious agent derived from streptococci and with “natural” influenza virus.

On intranasal inoculation of mice with pneumotropic streptococci, so far removed from original source as to exclude the possibility of the passive transfer of “natural” influenza virus, and on serial passage through mice and allantoic fluid of embryonated chicken eggs of emulsions or filtrates of emulsions of pneumonic lungs thus obtained, a pneumotropic, filtrable, transmissible, infectious agent has been obtained from each of twenty-nine cultures of streptococci. These twenty-nine cultures consisted of fourteen cultures from the nasopharynges or blood of thirteen persons having acute epidemic influenza, eight cultures of streptococci from a milk supply and two from a strain isolated from freshly fallen snow during an epidemic of influenza and five strains isolated by me from “natural” influenza virus that had been sent to me for study.

Control inoculations, made in 712 mice during the course of these experiments with (1) emulsions or filtrates of emulsions of lungs of normal mice and of the few uninoculated mice in which lesions of lungs had been found–lesions which, however, were clearly different from those in test mice–(2) filtrates of dextrose-brain broth cultures of the streptococcus and (3) filtrates of dextrose-brain broth and chick-embryo medium and inoculations made in 1,142 mice with non-pneumotropic streptococci from sources remote from influenza, did not yield the infectious agent.

The incidence, type and degree of gross and microscopic lesions that developed in the lungs of mice after intranasal inoculation of the infectious agent derived from pneumotropic streptococci after a number of serial passages were essentially the same as those of lesions that developed after intranasal inoculation of “natural” influenza virus. The incidence of isolation of streptococci from pneumonic lungs of mice that had received the experimental infectious agent also was similar to the incidence of isolation of streptococci from pneumonic lungs of mice that had received “natural” influenza virus. Isolations of streptococci and incidence of lesions, especially in the first number of serial passages, often ran parallel but, in general, isolations of streptococci diminished progressively with serial passages (figs. 1 to 7).

Strains of streptococci isolated from the pneumonic lungs of the two groups of mice, those receiving the experimental infectious agent and those receiving “natural” influenza virus, had moderate pneumotropic virulence and five strains of streptococci from the latter group, far removed from virus, yielded the infectious agent on successive passage of lung material, beginning with the streptococcus. The streptococci from both groups were agglutinated specifically by the influenza anti-streptococcic serum and by convalescent influenza serum.

The infectious agent produced from streptococci was as filtrable and as infective for embryonated chicken eggs as “natural” influenza virus and remained viable on preservation in 50 per cent glycerin for as long as three months.

The invasive power of both the experimental infectious agent and virus and the influenzal type of streptococcus was neutralized by the influenza anti-streptococcic serum and by convalescent influenza serum but not by normal horse serum and normal human serum.

Mice that were immunized intranasally or intraperitoneally with vaccines prepared from streptococci that had been freshly isolated from nasopharynges or blood of persons having symptoms of acute influenza became resistant to intranasal inoculation of the experimental infectious agent and “natural” influenza virus.

The neutralization of invasiveness (1) of the pneumotropic streptococci by treatment with influenza anti-streptococcic serum, (2) of the experimental infectious agent derived from streptococci and (3) of “natural” influenza virus by the influenza anti-streptococcic serum and influenza convalescent serum and the protection of mice against both the filtrable infectious agent derived from streptococci and “natural” influenza virus by intranasal and peritoneal immunization with the influenza streptococcic vaccine afford explanation, not heretofore available, for the highly favorable results that have been obtained by some with properly prepared streptococcic vaccines in prevention of epidemic influenza and related epidemic respiratory infections. Since the pneumotropic streptococcus, the filtrable infectious agent obtained from streptococci and “natural” influenza virus are antigenically related, protection should accrue alike from the use of properly prepared streptococcic and viral vaccines.

The possibility that the infectious agent obtained in these experiments might represent pickups of latent or spontaneous pneumotropic virus, described by others, in mouse stocks was considered and a report has been withheld until the evidence against such possibility seemed conclusive. Since the different strains of the infectious agent obtained by me from pneumotropic streptococci and “natural” influenza virus, isolated by others, are similar, the possibility of pickups of the latent virus in mice applies equally to my experiments and to the isolation and propagation in mice of “natural” influenza virus by others. The controls in both instances suffice, it would seem, to eliminate this possibility.

The data obtained in this study indicate (1) that the pneumotropic, filtrable, transmissible infectious agent obtained from pneumotropic streptococci appears to be true influenza virus, as now understood; (2) that pneumotropic streptococci in cases of influenza and related respiratory infections, such as the common cold, primary atypical pneumonia and influenzal bronchopneumonia, may be an important source of what is now considered virus; (3) that the virus may be the small, filtrable, highly invasive, relatively nonantigenic phase of the large, cultivable, highly toxic and antigenic pneumotropic streptococcus and (4) that cultivable, highly specific types of streptococci may play the primary and perhaps the major role in etiology, epidemiology and immunity in epidemic influenza and related diseases. It is hoped, through the publication of the results obtained in this broader approach than has yet been made to the puzzling problem of influenza and related epidemic respiratory infections, that the methods which have led to these results, or better ones embodying the essential principles, will be adopted by others.

References Cited:

1. Andrewes, C. H., Laidlaw, P. P., and Smith, Wilson: “Influenza: observations on the recovery of virus from man and on the antibody content of human sera.” J. Exp. Path., 16: 566, 1935.
2. Blanton, W. B., and Irons, E. E.: “A recent epidemic of acute respiratory infection at Camp Custer, Michigan.” A. M. A., 71: 1988, 1918.
3. Burnet, F. M., and Foley, M.: “Results of intranasal inoculation of modified and unmodified influenza virus strains in human volunteers.” J. Australia, 2: 655, 1940.
4. Evans, Alice C.: “Etiology of epidemic encephalitis; virulent bacteria cultivated from so-called herpetic and encephalitic viruses.” Health Rep., 42: 171, 1927.
5. Falk, I. S., Harrison, R. W., McKinney, R. A., and Stuppy, G. W.: “Experiments on the etiology of influenza; preliminary report.” A. M. A., 93: 2030, 1929.
6. Francis, Thomas, Jr. and Magill, T. P.: “Immunological studies with virus of influenza.” Exper. Med., 62: 505, 1935.
7. Henle, Werner, Henle, Gertrude, and Stokes, Joseph, Jr.: “Demonstration of the efficacy of vaccination against influenza type A by experimental infection of human beings.” Immunol., 46: 163, 1943.
8. Kendall, A. I.: “Observations upon the filterability of bacteria, including a filterable organism obtained from cases of influenza: Studies in bacterial metabolism, CI.” Science, 74: 129, 1931.
9. Long, P.H., Buss, E. A., and Carpenter, H. M.: “Etiology of influenza; transmission experiments in chimpanzees with filtered material derived from human influenza.” A. M. A., 97: 1122, 1931.
10. Rickard, E. R., Lennette, E. H., and Horsfall, F. L., Jr.: “A comprehensive study of influenza in a rural community.” Health. Rep., 56: 2164, 1940.
11. Rosenow, E. C.: “Studies in influenza and pneumonia: II. The experimental production of symptoms and lesions simulating those of influenza with streptococci isolated during the present pandemic.” A. M. A., 72: 1604, 1919.
12. Rosenow, E. C.: “Studies in influenza and pneumonia: III. The occurrence of a pandemic strain of streptococcus during the pandemic of influenza.” A. M. A., 72: 1608, 1919.
13. Rosenow, E. C.: “The relation of streptococci to the filtrable virus of epizootic encephalitis of the fox.” Infect. Dis., 48: 304, 1931.
14. Rosenow, E. C.: “Observations on filter-passing forms of Eberthella typhi (Bacillus typhosus) and of the streptococcus from poliomyelitis.” Staff Meet., Mayo Clin., 7: 408, 1932.
15. Rosenow, E. C.: “Observations with the Rife microscope of filter-passing forms of microorganisms.” Science, 76: 192, 1932.
16. Rosenow, E. C.: “Seasonal changes in the cataphoretic velocity and virulence of streptococci, as isolated from well persons, from persons having epidemic or other diseases and from raw milk.” Infect. Dis., 53: 1, 1933.
17. Rosenow, E. C.: “Cataphoresis as a control of specificity of streptococcal vaccines; influenzal streptococcus vaccine in the prevention and treatment of infections of the respiratory tract.” Immunol., 26: 401, 1934.
18. Rosenow, E. C.: “The relation of streptococci to the viruses of poliomyelitis and encephalitis; preliminary report.” Staff Meet., Mayo Clin., 10: 410, 1935.
19. Rosenow, E. C.: “Isolation of bacteria from virus and phage by a serial dilution method.” Path., 26: 70, 1938.
20. Rosenow, E. C.: “Microdiplococci infiltrates of natural and experimental poliomyelitic virus compared under the electron and light microscopes,” Staff Meet., Mayo Clin., 17: 99, 1942.
21. Rosenow, E. C.: “The relation of neurotropic streptococci to encephalitis and encephalitic virus.” Staff Meet., Mayo Clin., 17: 551, 1942.
22. Rosenow, E. C.: “Studies on the relation of a neurotropic streptococcus and virus to epizootic encephalitis of wild ducks.” Cornell Vet., 33: 277, 1943.
23. Rosenow, E. C.: “Poliomyelitis; the relation of neurotropic streptococci to epidemic and experimental poliomyelitis and poliomyelitis virus diagnostic serologic tests and serum treatment.” Bull. M. Research & Pub. Hyg., A44: 1, 1944.
24. Rosenow, E. C.: “The production of a filtrable infectious agent from alpha streptococci,” J. Clin. Path., 14: 150, 1944.
25. Rosenow, E. C.: “Isolation from milk supplies of specific types of green-producing (alpha) streptococci and their thermal death point in milk.” Minnesota Med., 27: 469, 550, 1944.
26. Rosenow, E. C., and Caldwell, H. W.: “Studies on the etiology and serum treatment of encephalitis during the epidemic in North Dakota and Minnesota (1941).” Int. Med., 17: 474, 1942.
27. Rosenow, E. C., and Caldwell, H. W.: “Epidemic encephalitis in North Dakota and Minnesota, 1941; studies on etiology, epidemiology and serum treatment.” Journal-Lancet, 63: 247, 1943.
28. Rosenow, E. C., Heilman, F. R., and Pettet, C. H.: “Observations on the epidemic of polio-encephalitis in Los Angeles, 1934.” Staff Meet., Mayo Clin., 9: 443, 1934.
29. Rosenow, E. C., Towne, E. B., and Wheeler, G. W.: “The etiology of epidemic poliomyelitis. Preliminary note.” A. M. A., 67: 1202, 1916.
30. Sharp, D. G., Taylor, A. R., Beard, Dorothy, and Beard, J. W.: “Morphology of the Eastern and Western strains of the virus of equine encephalomyelitis.” Path., 36: 167, 1943.
31. Smith, Wilson, Andrewes, C. H., and Laidlaw, P. P.: “Influenza: experiments on immunization of ferrets and mice.” J. Exper. Path., 16: 291, 1935.
32. Smorodintseff, A. A., Tushinsky, M. D., Droby-Shevskaya, A. I., Korovin, A. A., and Osetroff, A. I.: “Investigation on volunteers infected with the influenza virus.” J. M. Sc., 194: 159, 1937.
33. Taylor, A. R., Sharp, D. G., Beard, Dorothy, Dingle, J. H., and Feller, A. E.: “The isolation and characterization of influenza A virus (PR-8 strain).” Immunol., 47: 261, 1943.
34. Thomas, Lewis, Mirick, G. S., Curren, E. C., Ziegler, J. E., Jr., and Horsfall, F. L., Jr.: “Serological reactions with an indifferent streptococcus in primary atypical pneumonia.” Science, 98: 566, 1943.
35. Wiesner, R. R.: “Streptococcus pleomorphus und die sogenannte Spanische Grippe.” klin. Wchnschr., 31: 1101, 1918.