In 1887 Flügge responded to a call to Breslau (Wroclaw) entering arguably his most productive period. During the next 22-years Flügge developed a research institute devoted to studying infectious diseases and their transmission. He gathered a team of younger researchers who, under his direction, carried out interlocking investigations focussed on infectious diseases and their transmission, laying a foundation stone of our knowledge of aerogenous infection transmission (Fig. 1).

On Flügge’s arrival in Breslau the Hygiene Institute was housed in temporary accommodation, but new buildings, planned with Flügge’s assistance, were begun in 1897 and completed two years later. These comprised the Hygiene Institute, the Physiology Institute, a Pharmacology Institute and Clinic and numerous laboratories, a lecture hall and a library. Particularly noteworthy were the rabies and climatology stations [25].

Flügge’s 1897 article “Ueber Luftinfection” (Concerning aerogenous Infection) became his most frequently referenced work where he summarized conclusions regarding the existing literature and studies conducted under his direction in Breslau [1]. He collated evidence that a variety of infections can be transported following aerosolization by coughing or even just breathing or speaking or following aerosolization of infectious agents by air currents of varying speed from various surfaces.

Although TB studies carried out by Flügge and his assistants in Breslau have drawn the most attention it is notable that they comprise only approximately 40% of his academic publications during his 21-years in Breslau [26]. Other subjects studied included the dynamics of infectious diseases such as diphtheria, typhus, cholera and rabies; he was active in areas regarding health administration such as water hygiene, sterilisation of rooms or buildings and the overall living conditions of poorer sections of the population [25].

Carl Flügge in his office in Breslau approximately 1900 [27]

Robert Koch in his 1884 review of his TB research stated that there was no doubt that TB was a respiratory infection and that coughing by pulmonary TB (PTB) individuals loosened highly infectious sputum particles that if aerosolized could be inhaled, but he noted that these particles were big and unlikely to remain suspended in air for long [28]. This problem seemed resolved when George Cornet also demonstrated that Mtb-infected small dried sputum particles and fibres from infected clothing, bedding, handkerchiefs or carpets when aerosolized transmitted Mtb infection to guinea pigs [29]. Cornet demonstrated TB infected dust particles were frequently present in surroundings frequented by PTB patients.

Flügge’s 1897 article challenged this view of TB infection proposing that TB infection (and other infections) - could also result from inhalation of very small airborne, coughed droplets harbouring the infectious agent [1]. Early studies focussed on generation of airborne droplets during coughing and breathing. Microscope slides were extensively used placed at distances from patients varying from 50 cm to 150 cm usually at head height in front of, to the side or behind coughing or merely breathing patients, either vertical, horizontal or at a 45o angle assisting detection of smaller droplets.

The coughed droplet clouds and droplet features on microscopy were documented, in particular the contents, structure and size of droplets, the duration in the air and the microorganisms they carried. The cone shape of the coughed cloud demonstrated by the Tyndall effect was appreciated and slides were placed across the front of the cloud to study droplets at the clouds centre and periphery. Appropriate stains and microscopy enabled study of droplet contents, their structure and material impacted on slides could be washed off and cultured [30, 31].

TB related studies by Flügge’s colleagues and assistants under his mentorship in Breslau, were published in 1908 in a single volume “Die Verbreitungsweise und Bekämpfung der Tuberkulose” (The spread and containment of tuberculosis); 39 chapters dealt with investigations by Flügge’s colleagues and assistants into different aspects of TB, its spread and prevention. Highlights from these papers are briefly summarized in Table 2 [32]. Here Flügge introduced the term quantum designating the number of microorganisms necessary to establish infection anywhere or alternatively the volume of breathed air necessary to establish infection.

An unexpected and striking finding was that despite sputum microscopy and culture for Mtb being significantly positive, often only 40–50% or fewer PTB patients coughed droplets containing Mtb bacilli; examined repeatedly this percentage might increase. Results of individual patients varied considerably from day to day and early morning and winter were marked by a greater proportion of patients coughing droplets containing viable Mtb [30, 31, 33,34,35]. The great majority of droplets were found within a ½ meter of coughing patients and varied in size from 20 to 500 μm - containing varying numbers of bacilli often between 100 and 500; however, droplets furthest from the patient were usually smaller ranging in size from 20 to 30 μm and much less, containing very few bacilli. It was appreciated that the droplet diameter flat on microscope slides was at least three times greater than during flight or as hanging droplets; it was likely that these smallest droplets might remain airborne longer and be more suitable for inhalation [36]. Failing this they might settle on a variety of surfaces and after drying again become aerosolised by air currents caused by household activities.

For more sophisticated studies a sterilised glass chamber was constructed 3.2 cub meters in size within which a TB patient sat having put on a sterile overcoat and over-shoes and washed hands and face with sublimate. Following coughing or only normal respiratory movements by occupants, air aspirated from the chamber above head-height at predetermined speed and volume was passed through normal saline; following centrifugation the fluid sediment was cultured. Culture media or normal saline in settle-plates that could be opened and closed from outside the chamber at specific time points, were placed at different heights and distances from coughing patients within the chamber [31, 34]. Verification of Mtb by culture most often followed injection of concentrated specimens into a guinea pig’s abdomen. Settle plates above head height were opened 30-minutes or longer after a patient left the chamber. Similarly, air could be drawn continuously from the chamber and passed through normal saline before or after a patient left the chamber and the centrifuged sediment subject to guinea pig culture to demonstrate the airborne presence of viable Mtb.

Amongst the most influential early studies were Laschtschenko’s [34], demonstrating the wide aerogenous distribution of various microorganisms following coughing, sneezing or talking with varying vigour. After a researcher took fresh cultures of Bacillus prodigiosus (Serratia marscens) which produces the red dye prodigiosine into his mouth and spoke loudly or softly in a large lecture hall or carried out other manoeuvres a wide, prolonged distribution of micro-organisms occurred. B prodigiosus was documented at ceiling height at distances of 9-meters or more from source; these findings elicited widespread comment and concern. A similar much quoted study was later carried out in the British Houses of Parliament by Gordon, also demonstrating a similar disturbing wide distribution of B prodigiosus within the debating chamber (Anonymous 1906).

Regarding Mtb dissemination Laschtschenko assessed 21 ambulant PTB patients, not acutely ill, but only 4 (19%) coughed Mtb-containing droplets seen on microscopy of slides 50–100 cm in front of the patient. In further experiments 9 coughing patients were individually seated in the sterile chamber for 60–90 min, coughing when necessary. Five Petri dishes containing normal saline were placed at head-height or in the upper-most corners of the chamber at 90–170 cm distance from each patient. In 4 (44%) instances Mtb was identified in the Petri dishes by guinea pig infection. In later similar studies conducted by Bruno Heymann Mtb cultures were also positive when closed Petri dishes at ceiling height were opened 30 min and longer after subjects left the chamber. Further proof of the airborne presence of viable Mtb in chamber air was provided when air aspirated from the chamber for 60–90 min in the presence of PTB patients was passed through normal saline and Mtb bacilli cultured from the saline through which the chamber air was drawn [31]. Laschtschenko [34] also investigated 20 PTB patients at different disease stages during cough-free periods finding Mtb in oral secretions of 9 (45%), often in considerable numbers; he suggested that oral secretions with less mucinous sputum slime might be better suited for coughing of smallest aerogenous droplets. The motions of speaking, coughing or sneezing also generated droplets from oral fluids that could transport microorganisms.

Following Koch’s [28] and Cornet’s studies [29] and the assumption that TB transmission usually followed inhalation of airborne, dried sputum-dust particles and attached Mtb bacilli derived from coughed sputum particles lying on the floor or furniture or attached to clothing or bedding of active PTB individuals, early mass preventative measures aimed to ensure that sputum was collected in bowls and bottles, hygienically disposed, leaving no opportunity for it to dry and become airborne. This was also initially Flügge’s opinion. In the 3rd edition of “Grundriss der Hygiene” (1894) Flügge describes TB transmission as due to dried TB-infected sputum particles aerosolized by vigorous sweeping of carpets or equally vigorous shaking of blankets or clothing of PTB patients [43].

However, following studies of his colleagues and the publication of Flügge’s 1897 and 1899 papers Mtb infection following coughing and aerosolization of Mtb infected droplets could not be doubted, although the nature, duration and extent of danger remained uncertain. Paradoxically microscopy of glass slides assessing the numbers of droplets containing TB bacilli coughed by PTB patients who were either sputum smear-positive for acid-fast bacilli on microscopy or culture-positive for Mtb suggested that very sick patients with thick tenacious sputum might pose less risk of infection transmission than those less sick and ambulant whose muscular strength and thin fluid sputum might more easily allow aerosolisation of smallest airborne droplets carrying bacilli [2]; during several studies of droplets coughed by PTB patients, only approximately 40% were coughing Mtb bacill-containing droplets during a single evaluation [30, 35]. Coughing Mtb-containing droplets was also more frequent early mornings and during winter. Ziesché investigated 30 PTB patients with sputum microscopy-positive for acid-fast bacilli but only 12 (40%) coughed bacilli-containing droplets; this figure was similar to Heymann’s [30] findings. More frequent evaluation might increase the percentage of patients coughing bacilli.

Regarding potential inhalation of dried Mtb-infected sputum particles and their aerosolization Flügge later again revised his opinion following results of investigations by his colleagues [37,38,39, 42]. He grudgingly conceded that dried TB sputum dust particles might occasionally establish an intraperitoneal infection if injected into a guinea pig’s abdomen and if airborne might cause respiratory TB-infection in experimental animals [3]. Viable tubercle bacilli might also be attached to dried-sputum particles adhering to clothing, blankets and carpets [29]. Heymann then investigated dust from rooms frequented by phthisis patients, hospital wards and private homes finding viable Mtb in 18.4%, 24.3% and 12% of sites respectively [31]. Beninde [38] studied aerosolization of viable Mtb attached to sputum particles adherent to dried handkerchiefs. If the handkerchief carried relatively little sputum, remained unused and carried in a pocket drying for several days particles carrying viable Mtb bacilli could be aerosolised following very vigorous shaking. Successful Mtb transmission to guinea pigs was achieved in a minority of cases but under unnatural conditions. Even with complete drying of dust particles the chances of Mtb transmission were low. Earlier experiments probably failed because sputum specimens were not dry enough or the air speeds were too low. Only after complete drying and intense dispersion of infected material and delivering it by bellows directly into the guinea pig’s mouth did Cornet elicit experimental infection [29]; but any remaining dampness allowed conglomeration of dust particles inhibiting aerosolization [39].

Regarding the effects of drying on coughed droplets it is also relevant to recall the correspondence between Conrad Wissemann [44] and Flügge following Flügge’s landmark 1897 paper. Wissemann pointed out that in the absence of sufficient humidity droplet components such as mucin might dry and the droplets would shrink and could then be considered as “air dust” (Luftstäubchen); the smaller the droplets the more rapidly drying might occur. This could also be observed when droplets settled on glass slides and fluid rapidly evaporated. Wissemann suggested that in some instances Flȕgge’s droplet infection might be more accurately considered an “air-dust” infection.

Kӧhlisch’s animal experiments provided an interesting sidelight on inhalation of microorganism infected dust particles in comparison to that of infected mucous droplets [42]. Droplets of respiratory mucous penetrated much further into bronchi and bronchioles of guinea pigs than those associated with dust particles that appeared to cause considerable inflammation in bronchial and bronchiolar mucosa inhibiting further ingress of the relevant microorganisms.

In the light of these studies it is of considerable interest that using more sophisticated methodologies and culture of coughed aerosols from sputum microscopy-positive or culture-positive PTB patients very similar results were recently recorded with negative Mtb culture from aerosols coughed by approximately 40% of evaluated sputum microscopy and/or Mtb culture positive PTB patients. This confirmed again the divergence between positive sputum microscopy for acid-fast bacilli and/or Mtb culture results and the coughing of microscopy-positive droplets or aerosol-positive culture material [45,46,47,48]. As already described by Flügge’s colleagues more recent studies have also found that severely ill, bedridden patients were less likely to cough viable Mtb containing droplets [48, 49].

It was thus not surprising that attempts to infect guinea pigs housed in small boxes coughed at by unselected phthisis patients down relatively short tubes were frequently not successful in a majority of exposed guinea pigs [30, 35]. This cast doubts upon the reliability of the methodology. Heymann reported that of 25 guinea pigs exposed to coughing of sputum-microscopy Mtb culture-positive patients only 6 (24%) had post-mortem evidence of TB-infection; others also recorded similar results. [33]. When Heymann took more care selecting phthisis patients for infection-transmission studies he reported a significant increase in the percentage of exposed guinea pigs becoming Mtb-infected [30]. He also calculated that taking into account the quantum of air breathed by a guinea pig in comparison to that of a human 100 guinea pigs might be needed to match the respiratory capacity of adult human subjects.

In 1921 a final study of Mtb containing droplets coughed by PTB patients down a short tube towards guinea pigs housed in boxes was reported by Hippke [50]. Mtb-infection transmission to guinea pigs was studied in relation to coughed droplet size and numbers of bacilli coughed and Hippke reported that stable infection confirmed by post-mortem correlated with sputum microscopy findings of much smaller coughed droplets containing relatively few bacilli. Hippke noted that the smallest droplets might measure only 15–20 μm or less on microscope slides and carry very few bacilli even only a single bacillus. Hippke also determined whether patients coughed bronchial or oral droplets finding the former more efficient for infection transmission.

The nature of droplets coughed by PTB patients also elicited differing opinions. Heymann [30] described droplets as spherical, but oval at times with three concentric layers: an acellular outer coat, an intermediate collar of cellular material and cells and a central area of slime, threads of fibrin or leukocytes and in the middle frequently bacilli alone or gathered into colonies. Variants carrying slime flakes and a light slime cover were approximately 500 μm in diameter, containing more than 200 bacilli and might have arisen directly from a TB lung-focus. Considerably smaller droplets found on the furthest microscope slides consisted mainly of slime and carried leukocytes and scattered oral epithelial cells with considerably less extraneous material, probably aiding their flight, but frequently only 2–3 bacilli. Ziesché [35] considered that a simple division of droplet types could be oral and bronchial droplets; the latter often carrying surprising numbers of TB bacilli, but infrequently other microorganisms. The addition of saliva to these droplets was typical of oral droplets that he describes as often larger size, containing oral epithelial cells and frequently microorganisms such as streptococci and staphylococci in addition to occasional Mtb; these sank rapidly to ground compared to bronchial droplets that had greater “propulsive force”. He thus regarded oral droplets as relatively innocuous for TB spread carrying Mtb bacilli relatively infrequently and sinking more rapidly to ground.

However, Koeniger [40], like Laschtschenko [34], commented that considerable numbers of bacilli might be found in oropharyngeal fluids; the numbers “freed” from tenacious mucoid sputum, might depend upon the length of time spent in the oral cavity exposed to oral fluids and enzymes. Thus tenacious mucoid bronchial sputum might be less likely associated with infection transmission than more watery oral droplets carrying fewer bacilli.

Stimulated by the Breslau presence of Johann von Mickulicz-Radecki, known for promoting aseptic surgery, Flügge discussed surgical wound infections with him, but doubted that contact was a likely means of transmission during surgery [1, 51]. Also air currents within a surgical theatre probably had insufficient strength to aerosolize dried sputum particles carrying microorganisms. It was more likely that droplets aerosolized by speech, coughing or sneezing contributed to surgical wound-infection; further the presence of significant numbers of spectators, frequently speaking during surgery, also multiplied wound infection risks that would be undiminished even by considerable distances from patients as infected droplets could be carried by slightest air currents; patients themselves might be a source of wound-infection. Flügge’s discussions with Mickulicz contributed to the introduction of surgical masks for all present during surgery.