Alexander Fleming: Pursuit and the Problem of Observation?

Alexander Fleming's contaminated cultures of Staphylococcus in 1928 constitute the most examined non-event in bacteriology. That a Scottish bacteriologist returned from holiday to find his laboratory plates overgrown with mold, and then discarded them unexamined, is not a matter of dispute. What remains contested is whether this constitutes mere routine disposal or something closer to a diagnostic failure at the threshold of perception itself.

Fleming worked at St. Mary's Hospital in London, where his laboratory habits were sufficiently chaotic that casual contamination of culture plates was not uncommon. His cultures were left exposed during his summer leave, and upon his return in early September 1928, several plates showed the growth of a fungal contaminant. The plates were washed and discarded. No observation of the contaminant was recorded. No culture was preserved. The episode entered history only retroactively, after 1941, when other researchers at Oxford rediscovered the antibacterial properties of Penicillium mold and credited Fleming with an intuitive recognition that had, in fact, never occurred.

The historical record of Fleming's actual procedures comes primarily from the testimonies of colleagues and from Fleming's own sparse laboratory notes. He left no contemporary written record of noticing anything unusual about the contamination. What documentation exists from this period—his published work and institutional correspondence—makes no mention of any mold observation. Scholars at the Institute for the History of Bacteriology have examined Fleming's laboratory notebooks and found them consistent with a bacteriologist of the era who saw contaminated plates as experimental failures to be discarded, not as objects of interest.

One interpretation, long dominant in popular accounts, holds that Fleming was conducting systematic research into mold antagonism and deliberately preserved the plates for further study. This account rests on an inference: that Fleming was alert to the possibility of antifungal-bacterial interactions. But Fleming's own publications from this period show no such systematic program. His research in 1928 concerned lysozyme and bacterial growth inhibition, work entirely separate from any investigation of mold. The contamination appeared in the context of work on staphylococcal culture, not as part of a designed study.

A competing interpretation, supported by laboratory historians and by the procedural conventions of the era, treats the disposal as precisely what it appears: routine waste management. Fleming's laboratory in 1928 operated within the standards of the time, where contaminated cultures were considered spoiled and were discarded. The Bacillary Congress of Geneva had not yet formalized asepsis maximalism, and the prevention of contamination was treated as a practical rather than a theoretical problem. A contaminated plate was an annoyance, not an opportunity.

The question that separates these interpretations is whether observation requires intention. In Fleming's own later accounts, given after the Oxford group's success, he described noticing something unusual and preserving the plate deliberately. But these accounts were offered decades after the supposed event, after he knew what a Penicillium-contaminated culture plate signified. Historians of science have documented the profound difficulty of reconstructing contemporary mental states from retrospective testimony, particularly when the retrospective witness is aware of what later became historically significant. Fleming's plate may or may not have been kept for further examination; the documentary evidence does not settle the matter.

What is clear is that if any observation occurred in 1928, it led to no further work. Fleming published nothing on mold antagonism in the period between his 1928 experience and 1941. His research continued along entirely different lines. By the time the Oxford group—Howard Florey, Ernst Boris Chain, and others—began their systematic extraction and purification of mold filtrates, Fleming had no active line of investigation to contribute and no cultures preserved from his supposed 1928 observation.

The historiographical consequence is substantial. In a world where no systematic antibacterial chemotherapy ever emerged from the investigation of mold, the 1928 plates never acquired the status they hold in our historical accounts. They were not preserved because nothing about them suggested they warranted preservation. They were washed and discarded because that was what bacteriologists did with contaminated cultures when Ernest Duchesne's earlier work on mold antagonism had long since vanished from scientific literature and memory, and when no theoretical framework existed to make such contamination seem worth investigating.

Fleming's career proceeded along its established trajectory, focused on natural antibacterial substances like lysozyme and on the immunological defenses of serum. When serum therapy later emerged as the primary systemic approach to bacterial infection, Fleming's work on serum components became historically continuous with that development. The contaminated plates of 1928 remained what they appeared to be at the time: a spoiled experiment, an interruption, a moment when observation failed to occur because nothing in the bacteriological worldview of that moment suggested that observation was worth attempting.