Pouilly-le-Fort Anthrax Trial (1881)

Summary

In May 1881, Louis Pasteur accepted a public challenge from a skeptical veterinarian named Rossignol and staged a controlled trial on a farm at Pouilly-le-Fort, near Melun. Twenty-five sheep received two vaccinations with attenuated anthrax; twenty-five controls received none. On May 31, both groups were injected with virulent anthrax. By June 2, all twenty-five unvaccinated sheep were dead or dying. All twenty-five vaccinated sheep were alive and well. The result was reported around the world as a proof of the germ theory and of Pasteur’s vaccine. What the trial did not reveal — and what Pasteur himself concealed — was that the vaccine actually used had been prepared by a chemical method involving potassium bichromate, not by the oxygen-attenuation method Pasteur had publicly championed. This gap between public claim and private practice, documented by historian Gerald Geison from Pasteur’s laboratory notebooks, remains the central interpretive controversy in the history of the trial.

Background

Pasteur’s Attenuation Work

The conceptual foundation for Pouilly-le-Fort was built on Pasteur’s earlier demonstration that anthrax bacteridia were the cause of the disease. In serial dilution experiments, a drop of anthrax blood was cultured in sterilized urine, then a drop of that culture sown into a second vessel, repeated to the fortieth flask; each successive culture, however dilute, killed the inoculated animals — making it “absurd to imagine that the last virulence owed its power to a virulent agent existing in the original drop of blood.” (Vallery-Radot, René, 1928) This serial culture work established the microbial nature of anthrax independently of any animal body fluid, and was the direct experimental basis for Pasteur’s claim to be working from a proven germ theory.

From this foundation, Pasteur discovered that chicken cholera bacteria lost their virulence when cultures were left exposed to air for weeks and then revived — an accidental observation that he systematized into a general principle of attenuation. He found that by varying the interval of oxygen exposure, he could produce graduated degrees of virulence, from cultures that killed ten out of ten animals down to cultures that killed none but conferred immunity. (Vallery-Radot, René, 1928)

He announced the general principle in February 1880: attenuated cultures, when used as vaccines, reproduced their attenuated virulence faithfully through subsequent generations, and animals thus vaccinated resisted later challenge with virulent preparations. (Vallery-Radot, René, 1928) The theoretical framing was that oxygen — the attenuating agent — produced a biologically depleted organism that could no longer compete successfully within the host body.

For anthrax specifically, Pasteur and his collaborators Chamberland and Roux developed an attenuation method based on culturing the anthrax bacillus at 42-43°C in chicken broth, a temperature at which the bacterium could live and reproduce but could not form spores. After eight days at this temperature, the culture killed only four or five of ten sheep; after ten to twelve days, it killed none but conferred immunity against subsequent challenge. (Vallery-Radot, René, 1928) This thermal method was the one Pasteur described publicly.

The Competitive Context: Henri Toussaint

Pasteur’s urgency at Pouilly-le-Fort was not merely scientific. Henri Toussaint, a veterinarian at the Toulouse veterinary school, had announced an anthrax vaccine before Pasteur and had used potassium bichromate as an attenuating agent. Geison’s reading of the notebooks makes clear that Pasteur and his team were well aware of Toussaint’s work and had themselves experimented with the bichromate method — while Pasteur continued publicly to maintain that his own oxygen-attenuation approach was the correct one. (Geison, 1995)

The trial at Pouilly-le-Fort was arranged by Hippolyte Rossignol, a veterinarian skeptical of Pasteur’s vaccine claims, and was explicitly conceived as a public test that would settle the question before witnesses. For Pasteur, it was also an opportunity to establish his vaccine against Toussaint’s competing claim. (Geison, 1995)

The Trial (May–June 1881)

Protocol

The trial began on May 5, 1881. Twenty-five sheep received two vaccinations with attenuated anthrax preparations, spaced approximately two weeks apart. Twenty-five sheep and several cattle served as unvaccinated controls. On May 31, all animals in both groups received an injection of virulent anthrax. Witnesses, journalists, and farmers attended throughout. (Vallery-Radot, René, 1928)

The Result

Pasteur’s son-in-law René Vallery-Radot recorded the outcome with documentary precision: a telegram from Melun arrived before the official June 2 observation, reporting that eighteen unvaccinated sheep were already dead that morning and the rest dying. All twenty-five vaccinated sheep were in perfect health. When observers arrived, the carcasses of twenty-two unvaccinated animals lay side by side; two others breathed their last as they watched. One vaccinated ewe died subsequently on June 4, but post-mortem examination confirmed the cause was pregnancy complications, not anthrax. (Vallery-Radot, René, 1928)

The result was celebrated across France and internationally. Rossignol, who had arranged the trial as a skeptic, was converted. The trial entered popular and scientific memory as a definitive proof of the germ theory and of vaccination against a bacterial disease.

The Result and Its Reception

The Pouilly-le-Fort result arrived at a moment of significant scientific controversy over the germ theory. The demonstration that a bacterial disease could be prevented by prior inoculation with an attenuated preparation — and that the result was visible and countable — provided an argument against which critics struggled. Koch later acknowledged attenuation as “a discovery of the first order,” a significant modification of his earlier skepticism, though he remained doubtful about practical field results. (Vallery-Radot, René, 1928)

Vallery-Radot’s biography presents the trial as a culminating vindication: doubt converted to certainty, skeptics silenced, Pasteur’s method confirmed before the world. (Vallery-Radot, René, 1928) This account became the template for how the trial was taught for most of the twentieth century.

Geison’s Revisionist Reading

The Notebooks

Gerald Geison’s 1995 study The Private Science of Louis Pasteur drew on Pasteur’s laboratory notebooks — including the critical Cahier 91 — to present a different account of what actually happened at Pouilly-le-Fort. Folio 106 of Cahier 91 records Pasteur’s experiments with potassium bichromate as an attenuating agent. Folio 113 makes clear that the vaccine prepared for the public trial was the bichromate preparation — Chamberland’s method — not the oxygen-attenuation method Pasteur had publicly championed. (Geison, 1995)

The decision to use Chamberland’s method rather than his own was driven, Geison argues, by practical necessity combined with competitive urgency. The oxygen-attenuation method had not yet been refined to guarantee reliable results under field conditions. Pasteur needed to beat Toussaint to a successful public demonstration, and the bichromate method was more dependable. (Geison, 1995)

The Theoretical Stakes

The problem was not only one of attribution. The discrepancy between method and public claim bore on a fundamental theoretical question: did vaccines work by biological exhaustion of the microorganism’s virulence — Pasteur’s public theory — or by chemical attenuation, as the bichromate method implied? Pasteur’s commitment to the biological theory of immunity was contradicted by his private reliance on Chamberland’s chemical approach. He presented results that were genuine while misrepresenting the mechanism that produced them. (Geison, 1995)

Geison places this firmly in the category of ethically significant discrepancies between private and public science, distinguishing it from the routine rhetorical ordering of messy experimental data that characterizes all scientific publication. It was a deliberate public misrepresentation of the method used and of the theoretical basis for the result. (Geison, 1995)

Independent Confirmation

Two pieces of independent evidence corroborate Geison’s reading. First, Adrien Loir — Pasteur’s nephew and laboratory assistant — wrote in a manuscript of around 1937 (unpublished at the time Geison was working) that the vaccine used at Pouilly-le-Fort was indeed the bichromate preparation. His insider testimony was the first direct confirmation of what the notebooks showed. (Geison, 1995) Second, the Italian historian Antonio Cadeddu independently analyzed the Pasteur papers and reached the same conclusion through his own scholarship. (Geison, 1995)

Chamberland and Roux

The situation was further complicated by what it meant for Pasteur’s collaborators. Chamberland had developed the bichromate method and knew it was the preparation being used. Roux was aware as well. Yet the public credit for the vaccine, and for the theoretical explanation of its success, went entirely to Pasteur and to his oxygen-attenuation account. (Geison, 1995)

Historical Significance

The Pouilly-le-Fort trial matters in two distinct and separable ways. As a demonstration, it was real: the vaccinated sheep survived and the controls died, and this was observed by credible witnesses including skeptics. Anthrax vaccination worked, and Pasteur’s program of attenuating microbial virulence for immunological purposes had opened a new era in medicine.

As a historical episode, it presents a textbook case of the gap between private and public science. Pasteur was competing for priority, operating under field-condition pressures that made his preferred theoretical method unreliable, and publicly committed to a theory he could not immediately deliver on. The result was accurate; the description of how it was achieved was not.

Geison’s analysis raised questions that went beyond Pasteur to the broader practice of science: how much of the history of discovery consists of results being presented through frameworks that misrepresent the actual process that generated them? The Pouilly-le-Fort affair does not diminish the importance of the result, but it complicates any straightforward reading of it as a validation of a theoretical program.

Historiographical Debate

The Hagiographic Tradition

Vallery-Radot’s authorized biography of Pasteur — first published in French in 1900 — established the canonical narrative of the Pouilly-le-Fort trial: a moment of scientific triumph, a skeptic converted, a theory vindicated. (Vallery-Radot, René, 1928) His account of the June 2 result relies on documentary material including a telegram reading “stunning success” and careful descriptions of witnesses arriving to find carcasses and living sheep. Paul de Kruif’s Microbe Hunters (1926) popularized the same narrative internationally, cementing the trial’s place as a founding legend of bacteriology.

The key structural feature of the hagiographic account is that it treats the trial’s public framing as identical with its scientific content. Pasteur publicly claimed the vaccine was prepared by oxygen attenuation; the vaccine worked; therefore oxygen attenuation was validated. The notebooks that would have complicated this story were not publicly available, and Pasteur’s collaborators maintained silence during his lifetime.

Geison observes that Pasteur’s own published papers are an exemplary case of what Peter Medawar identified when he asked “Is the Scientific Paper a Fraud?” — not a fraud in the legal sense, but a retrospective reconstruction that presents the logic of the result as though it were the logic of the investigation. The vaccine paper at Pouilly-le-Fort went further: it misrepresented not just the reasoning but the method. (Geison, 1995)

Geison’s Critical Reading

The revisionist interpretation rests on the laboratory notebooks, and its scholarly authority depends on the quality of that evidence. Geison’s reading of Cahier 91 is supplemented by two independent pieces of corroboration — Adrien Loir’s 1937 manuscript and Antonio Cadeddu’s independent analysis — which reduce the risk that the notebook evidence is being over-interpreted. (Geison, 1995) (Geison, 1995) The convergence of three independent sources of evidence (notebook entries, insider testimony, and an external scholar’s independent analysis) makes the core claim that the bichromate method was used substantially more secure than any single source would have warranted. Geison further argues that Pasteur’s published accounts did not merely fail to disclose the true method but actively misrepresented it: from Pasteur’s public statements, one could never have guessed that the triumphant results at Pouilly-le-Fort had been achieved by a method notably different from the one to which he had so publicly committed himself.(Geison, 1995)

The more contested interpretive question is how to characterize what Pasteur did. Geison explicitly distinguishes his “ethically significant” category from both outright fraud (the vaccine did work) and the routine rhetorical packaging of scientific results. (Geison, 1995) The vaccines were effective; the sheep survived; anthrax vaccination was real. What was misrepresented was the mechanism — which method produced the attenuation and, by implication, which theory of immunity the result supported.

Theoretical Stakes of the Discrepancy

The most enduring historiographical question is whether the Pouilly-le-Fort misrepresentation mattered theoretically, or was merely a matter of credit and priority. Geison argues it mattered theoretically. (Geison, 1995) The distinction between biological attenuation (Pasteur’s public theory) and chemical attenuation (the bichromate method’s implication) bore on whether vaccines worked by depleting the pathogen’s capacity to compete within the host — Pasteur’s general theory of immunity — or by some chemical mechanism whose relationship to the organism’s biology was less clear.

By presenting a chemically attenuated vaccine as evidence for biological attenuation theory, Pasteur gave his theoretical framework an empirical support it had not earned. This meant that subsequent work proceeding on the basis of that theoretical validation was building on a false premise about what the Pouilly-le-Fort result had actually demonstrated.

[HUMAN NOTE]: None yet.

See Also

• Louis Pasteur
• germ-theory
• anthrax
• vaccination
• attenuation
• charles-chamberland
• discovery-of-vaccination-1796
• private-science

Sources

Evidence drawn from:

• Geison, G.L. (1995). The Private Science of Louis Pasteur. Princeton University Press. Ch. 6 — gei95-ch06-001 through gei95-ch06-010
• Vallery-Radot, R. (1928). The Life of Pasteur. Doubleday. Ch. 9, 10, 11 — vr28-ch09-003, vr28-ch10-001, vr28-ch10-002, vr28-ch10-004, vr28-ch10-005, vr28-ch11-006