The story of Napoleon's fatal illness has fascinated and divided historians, medical experts and even lay observers, for the best part of two centuries. It has always been a great temptation for the French to blame the British for the emperor's death, be it through a deliberate plot on the part of Hudson Lowe, or on account of the appalling living conditions Napoleon had to endure at Longwood House. Disagreement with this theory, on the other hand, would appear to have led others to interpret his death as a settling of old French scores. The time has come, therefore, to take an objective look at the facts and findings as they stand, taking care to steer clear of speculation, as difficult as this may be when dealing with the life and personality of the Emperor.
I. THE ILLNESS AND ITS BACKGROUND
In the five and half years that he spent on St. Helena, Napoleon did not, as is often supposed, contract a single and subsequently fatal illness; rather it was a series of ailments and conditions which caused his general state of health to deteriorate gradually. Although Napoleon's condition did worsen during the course of 1816 and the first half of 1817, it does not seem to have been any worse than that of his companions. As a result of the difficult climate, inadequate food resources and poor sanitary conditions on the island, many of the people arriving from Europe succumbed to serious ailments, particularly bowel disorders. The death rate among the ranks of the British army and navy was as high as 7% a year. General Gourgaud was the first to be affected although in his Journal (Diary) he does not give a detailed description. From May 1816 onwards, Napoleon began to consult the O'Meara for insomnia, head aches and “gouty” pains. This in turn took its toll on the Emperor's morale and he became melancholy and irritable. The periods of work and sickness alternated until September 1817, when the symptoms became more marked and Napoleon started to complain of pain on the right side of his abdomen. All witnesses reported significant swelling of the ankles and general weakness in his legs. O'Meara diagnosed hepatitis, administered calomel (a common medicine at the time made from mercury chloride) and began drafting reports; the sending of these to the governor greatly angered the Emperor. Hudson Lowe then came into direct conflict with O'Meara, accusing him of deliberately dramatising the situation in a bid to obtain Napoleon's repatriation on health grounds by ascribing the hepatitis to the living conditions on St. Helena. On 25 July 1818, Hudson Lowe managed to have O'Meara recalled to London.
On the night of 16 and 17 January 1819, Napoleon was taken so seriously ill that Bertrand and Montholon feared the worst. Dr John Stokoe was called in. His diagnosis of hepatitis diagnosis was much disputed and this criticism simply served to confirm the governor's suspicion that Napoleon's illness was in all probability a complete fabrication. Antommarchi, a Corsican doctor, who arrived at Longwood in September 1819, similarly diagnosed a liver disease and recommended that the Emperor should take more physical exercise. This he did and there was a marked improvement. However, he was not in remission for long and by the middle of 1820 the illness had begun to return. The horse ride he attempted in early October but could not finish, was to be his last. The illness was now advancing at any alarming rate, with Napoleon suffering pain on the right side of his abdomen, pain in his shoulder, fevers, coughs, chills in his legs, gingivitis, and alternating diarrhoea and constipation, etc. On New Year's Day 1821, it was clear to all the Emperor's closest companions that Napoleon would not live to see the next. On 17 March he was confined to his bed, where he remained, apart from a few occasions, until his death on 5 May. Antommarchi, whom the Emperor accused of not attending to his needs with sufficient care, prescribed tartar emetic (a vomiting agent containing antimony) which further weakened the patient. Napoleon then called for a drink made with bitter almonds, which should not have been administered. The English naval doctor Arnott was then called for. To begin with, he did not consider the illness to be that serious, but ultimately agreed that Napoleon was nearing the end. After consulting with two English colleagues and despite Antommarchi's opposition, Arnott prescribed a massive dose of calomel (10 grains compared with the normal one or two), causing severe haemorrhaging and death.
II. THE AUTOPSY
Antommarchi, assisted by seven British doctors, performed the autopsy at 2 pm on 6 May. The doctors were not, however, in agreement with Antommarchi's report and produced their own; Antommarchi then wrote a second report which differed slightly from the first. In one of his reports, Antommarchi stated that “the liver was swollen and of an unnaturally large size”; and in another that “the spleen and hardened liver were extremely large and swollen with blood; the liver tissue, red and brown in colour, showed no other significant alteration in structure. [ … ] the liver, which had suffered chronic hepatitis”. The organ that was found to be worst affected was the stomach. In addition to being heavily corroded with a “cancerous ulcer”, the stomach lining also contained a hole “some three lines in diameter” (7 mm). The autopsy also reported significant annular swelling near the pylorus muscle, which was described as a malignant cancerous growth. The discussion which ensued between the various doctors and with Hudson Lowe, Bertrand and Montholon, was extremely heated: ultimately the general consensus of the various reports and correspondence sent back to Europe centred on the “cancerous growth near the pylorus”; this conveniently had Napoleon dying of the same causes as his father and sister Élisa, thus suggesting hereditary predisposition. The British and French, having reached this compromise and being duly satisfied, met on 12 May at Plantation House for a reconciliatory tiffin, according to Major Gorrequer's Diary. Of course, once back in Europe, each member of the respective camps was to take the liberty of speaking his mind and giving his own opinions.
III. THE REACTIONS
For some one hundred and fifty years, the Emperor's illness continued to be the focus of wide-ranging speculation, some of it “extravagant and fantastic in the extreme” (Ganière). Indeed, what these interpretations did have in common was the fact that the illness diagnosed by the doctors examining the case was generally the one they specialised in. Another key source of controversy stems from a problem of language and usage: in 1821, the words ulcer and cancer were used to describe the same condition. For it is clear that, far from being emaciated, Napoleon had simply lost weight through several weeks of not eating, and did not die of cancer as we understand it today. Only the swelling around the pylorus muscle could be considered a developing tumour, its presence being connected with the effects of the earlier hepatitis condition and above all a serious corrosion of the stomach lining, leading to its perforation. For the lay observer, all of this may seem more than enough to explain Napoleon's death. For an explanation of why this stomach ulcer developed, we need only look at the psychosomatic studies conducted over the last few decades. “During periods of anxiety, inner turmoil and serious vexation, the stomach lining changes structure and swells. These findings show [ … ] that nervous dyspepsia can quite easily develop where worry, irritation, hostility and other difficulties of this nature are present, and that secondary organic changes, notably ulcers, can then occur; it is ulcers of this kind, when aggravated by constant nervous tension, which deteriorate until there is haemorrhaging and perforation of the stomach lining” (Gilbert Martineau). The Emperor's physical inactivity, imprisonment, continual ruminating over of the past and failed attempts to obtain repatriation – all factors present to a large extent in his exilic writings – are alone sufficient justification as to why the above process developed as it did.
IV. THE POISONING THESIS
During the 1960s, as the latest advances in medical research and psychology began to produce a reasonable and well-founded solution to the “Napoleon question”, Sten Forshufvud, a Swedish dentist and toxicologist, came to the conclusion, from a reading of Marchand's Mémoires published only in 1952 and 1955, that the symptoms and ailments related by Napoleon's first valet and in the other diaries from the period were consistent with those observed in cases of chronic arsenical intoxication. Forshufvud had also uncovered clues pointing towards Napoleon's arsenical intoxication well before St. Helena, in particular in Borodino, Leipzig and Waterloo. It is these conclusions of slow and gradual poisoning and his book Napoléon a-t-il été empoisonné ? (the question mark being for the sake of form only), published in France in 1961, that provoked such interest and which continue to be debated with such passion nearly forty years on.
Arsenic acts by lodging itself in the body's outer tissue, in particular the hair, as it is absorbed by the body, although it can also bind itself to the hair through simple physical contact. Using modern radioactivity analysis, it is possible to measure the average arsenic content of a strand of hair, as well as in specific segments, thus allowing the level of contamination to be traced in relation to the hair's growth (a rate of between 0.35 and 0.5 mm a day). Sten Forshufvud obtained strands of Napoleon's hair, about whose origin there is no doubt, and had tests carried out by Professor Hamilton Smith of the Glasgow University Forensic Science Laboratory. Soon thereafter he was joined by Ben Weider, a Canadian businessman with a passionate interest in Napoleonic history. The concentration of arsenic is determined by irradiating the hair sample with neutrons and analysing the gamma rays emitted. The readings themselves are expressed in ppm (parts per million). Although this method may appear straightforward, it is in fact extremely difficult to apply because of the hair's relatively low mass and the risks of the sample being polluted during handling. Interpreting the results is therefore a delicate task, especially given that the “normal” level of arsenic by current reckoning is 0.8 ppm, rising to as much as 2 or 3 ppm, and higher when the atmosphere is polluted; we can only speculate about what the standard level might have been on St. Helena in the early 19th century. It is true that arsenic is present naturally in many forms and can be found as a trace element or an impurity in a wide range of products, foodstuffs and medication, for example, when using basic chemical processes. To begin with, H. Smith discovered a high mean level of arsenic in the piece of Napoleon's hair shaved on 6 May 1821 (a lock belonging to Marchand and kept by the Musée de l'Armée), but he also found the same level (10 ppm) in a strand of Napoleon's hair given to the artist Isabey in 1805 (14 germinal, Year XIII). This result confirmed Forshufvud's theory that Napoleon suffered chronic arsenic poisoning during 1820 and 1821, and presumably in 1805 as well. His thesis became the focus of large-scale media attention which is still echoed in public opinion nearly forty years on. That the Emperor was the only person at Longwood House to display virtually all of the symptoms attributable to chronic arsenical intoxication led Forshufvud to believe the poisoning was a criminal act. After examining a range of possible scenarios, he concluded that the Comte de Montholon was the prime suspect, and ultimately the man responsible for the poisoning. Indeed, numerous facts about him – his reputation as a scheming, boastful character, an inveterate storyteller, even a liar, the fact that he was the main beneficiary of the Emperor's will, his greed, his liking for extravagance and the extremely negative portrait painted by Napoleonic historians – all make him, in the absence of evidence to the contrary, the ideal suspect.
Subsequently, Forshufvud, Smith and Weider continued their research in two directions. The initial focus was a fresh analysis of increasingly short strands of hair (millimetres long) in order to establish a link between the rate at which the arsenic levels changed and the development of the symptoms described by eyewitnesses. The results were extremely varied, with readings of as much as 50 ppm locally for the hair shaved on 6 May 1821. Work then concentrated on all the elements of the story where Montholon's version of events, as contained in his Récits de la captivité, published in 1847, did not concur with the account of the other diarists on St. Helena. Indeed, as a direct result of the character of the work and the circumstances under which it was produced, the differences are legion (see Hélène Michaud). There is nothing inherently surprising about the fact that Montholon's Récits take such liberties. After all, he hoped to have them serialised in the popular newspaper La Presse; and Alexandre Dumas (père) also went to the fortress of Ham towards the end of 1844 and early 1845 – at the request of Louis Napoleon who sought to use the Récits to make his uncle's death that of a martyr and thus set the stage for his own future – with the intention of reviewing and correcting Montholon's manuscript as he saw fit. Polemicists of the day also seized on the story, with Le Globe, a rival publication of La Presse, commenting ironically: “Mr de Montholon's memoirs will be nothing more than the Memoirs of Mr Alexandre Dumas, who would have us see the Emperor in the same historical light as trout fishing antics and bear beefsteak” (a reference to Dumas' extraordinary accounts of his travels in Switzerland!). However, taking Montholon's Récits at face value, Forshufvud, Weider and Smith presented their findings in Assassination at St. Helena, a work published in 1978 in the United States and Canada, followed by The Murder of Napoleon by Ben Weider and David Hapgood in 1982. These works convict Charles de Montholon of Napoleon's assassination by chronic arsenic intoxication using the “Brinvilliers” method. The motive put forward by the authors is based on the theory that Montholon was a royalist agent and that the murder was ordered by the Comte d'Artois. A number of doctor historians, notably Paul Ganière and Guy Godlweski, have stated themselves sceptical of the poisoning thesis, and many historians on the other hand find the suggested motive as fundamentally implausible, particularly given the unequivocal Bonapartist views which Montholon continued to defend up to the time of the Second Republic.
Professor René Maury rekindled the debate with his publication of L'Assassin de Napoléon, ou le Mystère de Sainte-Hélène in 1994. In it, he reexamines, this time in greater detail, the correlation between the symptoms of arsenical intoxication and specific events in Napoleon's life on St. Helena, claiming to have pinpointed the vehicle for the poison – the vin de Constance which only the Emperor drank – and placing responsibility for the murder squarely on Montholon, a depraved individual driven by vengeance, jealousy and greed: vengeance because of the discharge from his post in 1812, jealousy because of the liaison between Napoleon and Albine de Montholon, and greed as demonstrated in the events surrounding the Emperor's will. Although seen more as a 'whodunnit' than a serious historical work, the book nevertheless served to resuscitate the controversy. In response to Maury's work, Ben Weider produced a revised edition of his 1978 book, Assassination at St. Helena revisited (1995), and published a new volume entitled Napoléon, Liberté – Égalité – Fraternité (1997). In both these works, he continues to accuse Montholon and maintains the theory that he was the Comte d'Artois' accomplice. He also released new analyses, this time carried out by the FBI, showing high levels of arsenic in strands of Napoleon's hair cut in 1816.
In July 1998, Professor J. Thomas Hindmarsh, pathologist and toxicologist of the University of Ottawa, and Doctor Philip F. Corso, of Yale University, published a remarkable study in the Oxford University Review entitled The Death of Napoleon Bonaparte: A Critical Review of the Cause. By diagnosing a gastric carcinoma, a tumour of the epithelium (as opposed to a sarcoma) which had developed on an earlier perforating ulcer, its authors sought, from a clinical standpoint, to settle the question of whether Napoleon had suffered a stomach ulcer or stomach cancer. As the carcinoma had not reached its terminal stage, the immediate cause of death must stem from the internal haemorrhaging brought on by the massive dose of calomel prescribed by the three British doctors. In toxicological terms, they set out to place in perspective the relative importance of the symptoms of arsenical intoxication displayed by Napoleon and put forward a range of supporting hypotheses, using as a basis reputed scientific publications. Unconvinced by this research, Ben Weider published another book in April 1999, this time in French, entitled Napoléon est-il mort empoisonné?. In it, he again develops Forshufvud's poisoning thesis together with the theory that the British and the French (specifically the Comte d'Artois) conspired to bring an end to Napoleon's life by using General Montholon as executing officer and the so-called “Marquise de Brinvilliers” method. Although better argued than earlier publications by the same author, the book nonetheless draws on some rather unsound historical interpretations and numerous quotations, sometimes taken out of context.
What conclusions can be drawn then, after forty years of theorising and debate? One thing is for sure: the presence of a significant level of arsenic in the Emperor's hair – further confirmed by analysis carried out in 1995 at the Saclay Nuclear Research Centre in France and at the University of Toronto -, is irrefutable. The vast majority of the issues raised by this arsenic presence were covered in Hindmarsh and Corso's study:
1) Post mortem explanation. Arsenic taken orally, like that absorbed through physical contact, collects on the outer wall of the hair. Contrary, therefore, to what Hamilton Smith claims, the precise origin, whether exogenous or endogenous, of the arsenic detected cannot be established through analysis. Indeed, though the fashion for powdered wigs and hair more or less ended with the Revolution, the practice of using powders made with arsenic oxide to treat hair was still common in the early 19th century. Certain sources from the period state that it was customary to treat with such powder hair cut for use as relics in order to preserve it and prevent parasite damage. The advantage with this theory, which to a large extent puts paid to any further debate, is that it would explain the high levels of arsenic recorded in the 1805 and 1815 strands of Napoleon's hair, both periods when poisoning would seem highly unlikely.
2) There is, however, good reason, given the clinical symptoms noted, some of them consistent with those of chronic arsenical intoxication, to examine the theory that intoxication of this kind, whether caused by internal or external factors, might have taken place.
– External origins. Although never established as being the case, the theory that arsenic poisoning was made possible though the dyes present in the new wallpaper used at Longwood House in 1819 – put forward by Jones and Ledingham in the scientific journal Nature (Oct. 1982) – cannot be ruled out altogether. There is also the possibility that the smoke from burning coal led to pollution of the atmosphere in the Emperor's small apartments. Hypersensitivity to arsenic on Napoleon's part would also explain why he displayed such marked symptoms.
– Internal origins. Before tackling the theory that the ingestion was a criminal act, the possible causes of accidental contamination need to be examined. As O'Meara and Antommarchi's diaries show, both prescribed pills and potions of various kinds, despite Napoleon's strong aversion to medication. These medicines, mostly prepared using very basic methods, could contain arsenical impurities. The vin de Constance, brought in specially from Capetown and reserved for the Emperor's exclusive consumption might also have been contaminated during its production or while being shipped. Also, the body of Napoleon, when exhumed in 1840, was found to be extremely well preserved, a factor which would further confirm the arsenical intoxication; however, the body had in fact been purified internally after the autopsy with eau de Cologne, a process which may have helped preserve it. Unfortunately, as there are no comparative elements, in other words analyses of hair strands from other people living at Longwood in the 1820s, it is impossible to tell whether or not the levels of arsenic measured differ from the normal levels for the same place and time. As Hindmarsh and Corso point out, the combined effect of the various causes may have led to the levels of contamination recorded.
In clinical terms, Forshufvud and Weider set out to establish a link between how the arsenic levels developed from January to May 1821 and how the Emperor's health fluctuated over the course of this period. The study, which is based on detailed hypotheses concerning hair growth rates and on analyses of 1-mm segments, reaches a number of disturbing conclusions, despite the inaccuracy of the analysis method (estimated by staff of the CEN at Saclay to be + or – 20% on 3-mm segments). Moreover, Hindmarsh and Corso point out that according to some researchers, arsenic, rather than remaining at the point where the hair has grown, in fact spreads throughout the strand by capillarity, thus making this results obtained with this method far less reliable. Forshufvud and Weider have also drawn up a list of thirty-two symptoms that indicate chronic arsenical intoxication, despite that fact that most of them are too general to be significant: headaches, fatigue, insomnia, depression, weakness of the legs, cough, sweating, tachycardia, irregular pulse, etc. If observation is confined to those symptoms regarded as the most specific (damage to the nerve centre and sensory organs, photophobia, hair loss, sensation of fever, chilled feet, spasms, agitation, etc.), it is clear, as Hindmarsh and Corso also demonstrate, that the most characteristic symptoms are not present: melanoderma, painful polyneuritis of the legs, keratisation of the skin, or Mees' lines on the nails. This finding, though not disproving the arsenic theory outright, does serve to put it in perspective. In response, supporters of the thesis that Napoleon was poisoned say that no one case ever displays all the typical characteristics and that the crime is rarely discovered as a result of the various illnesses which develop in its wake. While it is true to say that this ingenious, though not entirely watertight argument, does carry some weight, given the quarrelling of doctors for the last century and a half over the exact nature of the Emperor's illness(es), the fact remains that none of the speculation about criminal intoxication to emerge over the past forty years stands up to scientific scrutiny. Thus Professor Henri Griffon, Director of the Paris Police Department's Toxicology Laboratory – who agreed in 1961 to write a highly favourable preface to Sten Forshufvud's first book -, subsequently expressed doubts and distanced himself from the poisoning theory in an article published in the journal Historia in 1982. The works of Forshufvud and Weider, except for his most recent publication, which comes across as well-argued compared with earlier writings, seem essentially to take evidence that is merely disturbing and use it to back up the writers' private theory; through the persuasive powers of René Maury, on the other hand, the reader is carried along in a clever and convincing display of literary talent. Nonetheless the fact remains that the vast majority of the theories advanced by these authors are biased in a bid to achieve one common goal: to satisfy people's liking for mystery and historical enigmas. After all, imagination and fantasy have always played a key role in Napoleonic Legend, have they not? In all events, the “poisoner” theory is slightly less far-fetched than that of Georges Restif de La Bretonne who, in 1969, in Anglais, rendez-nous Napoléon, claimed that the body of Napoleon was exchanged for that of Cipriani, the Emperor's major domo and spy!
On a purely historical and medical level, the following combination (consistent specifically with the study by Hindmarsh and Corso): earlier chronic hepatitis + stress-induced stomach ulcer, developing eventually into perforation + deterioration of the ulcer to become carcinoma + a worsening of the overall condition by a dangerous cocktail (antimony and mercury), is more than sufficient to explain the Emperor's death, without considering the possibility that he was poisoned, or attacking the memory of Montholon, about whom the only positive aspect, in his troubled existence, was without doubt his devotion to the deposed Emperor, a filial devotion to which Napoleon's will bears witness. Equally, if we are to find any consensus of opinion in the various theories put forward, the scenario whereby the high ambient level of arsenical intoxication may have adversely affected a body already weakened by illness cannot be ruled out.
Perhaps in the end the root the cause of the Emperor's illness and subsequent death was his suicidal attitude at Longwood House and the unavailability of counselling (not yet invented), both of which could have helped him come to terms with his lost status as Head of State, to overcome the trials and tribulations of exile and see his future in a more positive light. What if he had agreed to negotiate the terms of his detention on the island, to examine more carefully Hudson Lowe's proposals, not all of which were malicious, to take an active part in social life on St. Helena by replying to invitations and by continuing to receive guests; what if he had refrained from turning over relentlessly in his mind those glorious victories of the past, if he had not constantly sought the reasons for his failures; what if he had not unwittingly promoted conflict within his entourage (officers, their wives and servants); what if he had taken regular exercise, refused to lead such a confined existence, not worked nights and slept during the day and eaten his meals in less haste…, could his life at Longwood have been, if not exactly enjoyable, then at least bearable. He would not have passed his days “eating himself up inside” and not have damaged his health irredeemably to die barely fifty years old. At the end of 1820, he would have moved into the new and well-appointed house built specially for him, to wait for adversaries to grow weary or fade away and for France's need for aid in time of crisis. In 1830, Napoleon would have been only 61 … General de Gaulle was 67 in 1958.
Additional bibliographical elements
Dictionnaire Napoléon, ed. J. Tulard, Paris: Editions Fayard, 1999, 2 vol. Entry: 'Empoisonnement de Napoléon', Jacques Macé, vol. 1, p. 720-724, bibliography
Jean Tulard, Napoléon ou le mythe du Sauveur, Paris: Editions Fayard, 1998,
new edition reworked and extended., 512 p.
Sten Forshufvud, Napoléon a-t-il été empoisonné?, Paris: Editions Plon, 1961
Ben Weider, Napoléon est-il mort empoisonné?, Paris: Editions Pygmalion – Gérard Watelet, 1999, 335 p.
Jacques Jourquin, Jacques Macé: 'L'affaire Montholon', Revue du Souvenir Napoléonien, n° 419, July-August 1998, p. 4-11
Jacques Macé, 'Le taux d'arsenic dans les cheveux de l'Empereur', Revue du Souvenir Napoléonien, n° 421, December 1998 – January 1999, p. 66-67
J. T. Hindmarsh, P. F. Corso, 'The death of Napoleon Bonaparte: a critical review of the cause', Journal of the History of Medicine & Allied Sciences, 53 (3), July 1998, p. 201-218
Jean-Claude Damamme, 'Napoléon a-t-il été empoisonné?' 1: for / Thierry Lentz: 'Napoléon a-t-il été empoisonné?' 2: against, Napoléon Ier – Le magazine du Consulat et de l'Empire, n° 3, July – August 2000, Dossier, pp. 34-36 and 37-41