“I cannot be sure that Forshufvud and Weider are right, but to prove them wrong, their opponent will have to produce an impressive hat and hope there is a rabbit in it” – Newsweek magazine, book review of The Murder of Napoleon by Ben Weider.
The debate amongst historians over the cause of Napoleon’s death dates from 1840, the year when his remains were exhumed (at 11am on 16 October) and repatriated to France from the island of St Helena. The controversy reached its peak at the time of the famous “Debate of the Century” organised by the Napoleonic Society of America on 11 September, 1994, and held in Chicago. Was Napoleon poisoned with arsenic by the Count de Montholon, his chamberlain, or did he die from cancer of the stomach? This debate has been dividing academic Napoleonic circles for 150 years and continues to provoke heated discussions.
Firstly, the inference of arsenic poisoning as revealed in the body’s excellent state of preservation – at the time of exhumation, Napoleon’s body was found to be intact, and arsenic acts as a preserving agent (it is still used today by taxidermists) – would seem to be refuted by the fact that the body of Henry IV, torn from his tomb by the revolutionaries, was found to be similarly intact. Just as – it is interesting to note – the body of Albine de Montholon, Napoleon’s mistress on St Helena, exhumed on the initiative of her daughter Napoleone. According to recent expertise, the preservation of the body is due principally to the dehydration resulting from the diarrhea brought on by arsenic. The preservative capacity of arsenic itself remains as yet unproven.
New analyses of Napoleon’s hair behove the same caution, since the presence of arsenic in the Emperor’s hair has not been found to be definitively conclusive. This presence could be due to contamination from external sources and therefore, not criminal. According to the chemist François Raspail, arsenic is omnipresent insofar as it infiltrates naturally the living body. In 1841, Ofila also demonstrated that arsenic is found consistently in human tissues.
Doubt also exists over the motive for the crime, if crime there be. Some suggest that Napoleon had a child by Montholon’s wife, which, needless to say, infuriated Montholon. Others subscribe to the idea that Albine de Montholon was certainly the Emperor’s mistress with the consent of her husband, but that this liaison dangereuse did not produce any offspring.
Both the proponents of the poisoning theory and those opposed to it put forward puzzling arguments. This paper seeks to settle the controversy by bringing current medical knowledge to bear on the problem.
The first paper to launch the theory that Napoleon was poisoned by arsenic appeared in the journal Nature on 14 October, 1961. At the same time, it refuted the generally accepted view that Napoleon’s death was due to cancer of the stomach. The poisoning theory was corroborated in the same article by a review of Napoleon’s medical history during the time he lived on St Helena, and an assertion that it was a huge dose of calomel administered to the Emperor, which provided the culmination to long-term arsenic poisoning. Calomel is Mercuric Chloride, a white powder used at that time as a purgative and intestinal antiseptic. But above all it was an analysis of a sample of Napoleon’s hair that prompted the poisoning theory, because it revealed a significant level of arsenic, which could have been given to Napoleon with criminal intent. In order to spare the audience the repetition of well-known facts, I refer those interested to the correspondence which followed the appearance of the article in Nature and which was published in Lancet in 1962.
I do not intend to comment, once again, on the results of the analyses mentioned above, which were carried out by the Department of Forensic Medicine in Glasgow using advanced neutron technology (perfected in the 70s.). I can only accept the interpretation of these analyses made by Jacques Macé and published last year. Nor am I in any way contesting the authenticity of the analysed sample of hair taken the day after Napoleon’s death, on 6 May, 1821, and now part of the private collection of Ben Weider. It is well known that the murder theory was first introduced by Dr Sten Forshufvud in his book Who killed Napoleon? It was subsequently taken up by Ben Weider who, arriving independently at the same conclusions, published them in a work written together with Forshufvud, The Assassination at St Helena. It seems to me significant that Sir Russell Brock, in his review of Forshufvud’s book, did not find his arguments convincing. Weider’s book also unleashed a torrent of emotions and doubts, but I do not intend to fan the flames of that debate. I should also mention Ben Weider’s most recent book, Was Napoleon poisoned?, published in 1999.
It seems unnecessary to repeat for the umpteenth time the arguments and counter-arguments used by those for and against the murder theory. Nor shall I come back once again to the well-known results of the hair analyses which have been used, wrongly in my opinion, to support the poisoning theory. Several observations speak against it. The characteristic clinical symptoms of chronic arsenic poisoning were absent, symptoms such as extreme weight loss, loss of nails, fatty degeneration of the liver (steatosis), skin pigmentation, patchy bleeding visible as dark red spots on the inner surface of the stomach, Mees’ lines and neurological symptoms. Furthermore, prolonged administration of arsenic leads to renal failure and anuria (that is, failure to produce urine). None of this was reported with respect to Napoleon and no physician, including those present at the autopsy, found any evidence of it. Also not reported was any evidence of vasodilatation (that is, reddening), a characteristic sign of the effects of arsenic on blood vessels.
Other characteristic signs of chronic arsenic poisoning were considered absent at the autopsy, notably hyperkeratosis, that is, thickening of the scaly layer of the skin on the hands and feet. As to the general lack of body hair noted at the autopsy, it is true that this is characteristic feature of arsenic poisoning, but it could equally be explained by a hormonal syndrome diagnosed in Napoleon.
In view of all this, I thought it proper to consider again the records of the autopsies performed twenty hours after Napoleon’s death, one of which by Francesco Antommarchi, Napoleon’s official doctor and a compatriot from Corsica. This report was published in his work The last moments of Napoleon. It is important to acknowledge the excellent reputation enjoyed by Antommarchi – in spite of his young age, 29 – as an anatomist and pathologist; he was well known in Italy. He stayed at Napoleon’s side on St Helena until his death at 5.49 p.m. on 5 May, 1821.
Despite occasional problems beginning in the summer of 1817, Napoleon’s health up until mid-September 1820 seemed relatively good. He did however suffer periodic episodes of intense weakness and it is fair to say that he had health problems throughout his life. Indeed, even on this point historians disagree. Starting on 18 September, 1820, more obvious symptoms began to appear, and from this date the development of a single specific illness – punctuated by periods of relapse and remission – can be observed affecting Napoleon. I have studied the accounts left by Napoleon’s companions on St Helena: Count Charles Tristan de Montholon, Marquis Emmanuel Las Cases, Baron Gaspard Gourgaud, Dr Barry O’Meara, Dr Francesco Antommarchi, Count Henri Gratien Bertrand, Louis Marchand (senior chamber valet of Napoleon for 10 years), and Drs Henry and Stokoe. They all agree. Napoleon suffered stomach troubles, pain in the upper part of the stomach, anorexia, nausea, vomiting, hiccup (irritation of the diaphragm), dysuria, lethargy, spikes of fever, diarrhea, constipation, abdominal cramps, excessive weakness, heavy perspiration and circulatory problems. Indeed, not much medical confidentiality here!!
In April 1821, Napoleon’s health deteriorated considerably and Dr Arnott, a British military surgeon called to assist, took over the management of the Emperor, until then under the care of Dr Antommarchi. On 3 May, despite Antommarchi’s objections, Arnott administered to the patient 64.8 centigrams of calomel, a dose 5 times greater than normal, prescribed jointly with Drs Shortt and Mitchell. Five or six hours later the patient had a violent response: persistent internal hemorrhage occurred, manifested by black vomit, passage of tarry stools, heavy perspiration and a rapid pulse. Death followed two days later.
Knowing that his father, Charles Bonaparte, had died of cancer of the stomach (he died in a cachectic state on 24 February, 1785 at the age of 39), and convinced that he suffered from the same disease, Napoleon requested an autopsy to be performed after his death. He wanted to make his son aware of this predisposition.
That Napoleon’s father had stomach cancer is generally accepted, but those who support the poisoning theory have cast doubts on this assumption. Furthermore, Weider refers to a work by the Ardmore Center for Research in Pennsylvania, according to which stomach cancer is not hereditary, and to a French doctor, Michel Ibos, who also postulates that stomach cancer is not hereditary. However, according to recent statistics, hereditary factors do play some role in stomach cancer, although their influence is not great. It has been established that the risk of developing stomach cancer is three times higher in families where one or more members have suffered from this disease. Types of food preservation and the environment also play significant roles in the formation of stomach cancer. They both add to the genetic factor and exercise their influence from an early age. It would be interesting to examine statistics showing the incidence of different types of cancer in Corsica. They do not exist, however. I have, on the other hand, been able to examine statistics listing the incidence of various cancers in different countries. Interestingly, in Italy, the incidence of stomach cancer is twice that in France. It is worth remembering that Napoleon was born in Corsica.
The autopsy report by Antommarchi agreed – except for one detail – with the observations made independently by the 17 persons present, eight of whom were doctors: Antommarchi, Henry, Shortt, Burton, Livingstone, Mitchell, Rutledge and Arnott. They all diagnosed stomach cancer complicated by a chronic perforation and hemorrhage, and associated with chronic gastric inflammation. When reading Antommarchi’s report, I did however notice a contradiction in his description of the liver. Having first stated that the liver showed no visible abnormality, he went on to say several lines later that it was affected by chronic hepatitis. Such contradictions are open to numerous interpretations. Paradoxically, Antommarchi was later to say that Napoleon died of chronic inflammation of the stomach and of acute hepatitis (inflammation of the liver) aggravated by the climate on St Helena. Antommarchi strongly argued in favour of the theory of hepatitis generated by the local climate, already long before Napoleon’s death. Even after his return from the island, Antommarchi continued to persist in this diagnosis, in spite of the absence of any obvious abnormality of the liver, as stated by himself at the autopsy, except its somewhat large size. Some hidden political influence has been suspected in this matter. Enlargement of the liver – due to fatty change – is one of the characteristic effects of arsenic poisoning. At autopsy, Napoleon’s liver was seen to be large and healthy. Large, but far from being pathologically enlarged.
The autopsy performed on Napoleon’s body provided other medical messages, but it is important to stress that, at that time, the autopsy could not benefit from microscopic examination of tissues, a procedure as yet inexistent. The descriptions were based exclusively on the appearance of the body and of the organs to the naked eye. One detail in particular caught my attention during my work on this study. The proponents of the poisoning theory, particularly Ben Weider, claim that Napoleon did not manifest any evidence of cancer. They emphasize the severe weight loss typically caused by stomach cancer and its tendency to spread to other parts of the body. Ben Weider points to the fact that Napoleon, during his illness, did not feel any tumor in his stomach. Napoleon however could not have felt the cancer in his stomach, firstly because of the position of the part of the stomach affected by the disease – this part was covered by the liver; and this part which was covered by the liver, and secondly because of the diffuse variant of his cancer.
In order to make my contribution to the debate, not only have I studied, once again, the autopsy records in question, but I have also reconsidered all the symptoms that Napoleon experienced in the period from his arrival on St Helena (on board the Northumberland, with 600 books in his trunks) in mid October 1815 to his death. In the light of this new examination, I suggest that Napoleon did indeed suffer from cancer and that he died of complications of the cancer, precipitated by a deadly dose of calomel.
The administration of calomel (a prescription which led unintentionally but directly to his death) two days before Napoleon’s death was a medical mistake and not a criminal act. Therefore, the accidental killers were the doctors Arnott, Shortt and Mitchell, but had they not made this professional mistake – no doubt, in good faith! -, and taking into account the autopsy findings and current medical knowledge, Napoleon would have died of his cancer within some 6 months. I believe this to be the case despite categorical statements by B.Weider claiming that the autopsy showed “absolutely no evidence” for the diagnosis of cancer.
Ben Weider bases this surprising statement on the absence of any cancerous tumor mass in the stomach at autopsy. The extensive pathological process involving the inner surface of the stomach, interpreted by Antommarchi as cancerous ulceration and scirrhous thickening, is explained by B.Weider as an effect produced by the calomel. However, stomach cancer in many cases does not take the form of a tumor mass, but involves the stomach more diffusely, producing thickening of the wall and ulceration of the inner surface, just as Antommarchi described. Chemical damage would have affected the whole lining of the stomach and the lower oesophagus, but such was not the case. As parts of the stomach, particularly the opening of the gullet and the greater curve were intact (cancer preferentially involves the lesser curve), I conclude that this picture better corresponds with a diffuse form of cancer than with chemical corrosion. Acute arsenic poisoning typically affects the stomach lining uniformly, causing swelling and – most typically – patchy bleeding seen as dark spots. Such a pattern was not noted at the autopsy.
There is a possible explanation for the serious deterioration in Napoleon’s health a few days before his death. It has not been pointed out that the cancer-related perforation of the stomach, although sealed by an inflammatory reaction which pushed the stomach against the liver, could have caused the intense abdominal pain which we know Napoleon suffered. I believe that it was this condition which provoked the critical deterioration of Napoleon’s health. A perforation was reported at the autopsy. In Forshufvud’s opinion, the excessive dose of calomel given to Napoleon 48 hours before his death could explain the pathological findings in the stomach. I doubt however that the perforation was that recent because, at the autopsy, not only was the perforation completely sealed, but it was also firmly attached to the surface of the liver. This suggests that the perforation occurred prior to the absorption of calomel.
The frequent illnesses from which Napoleon had suffered since 1815 intensified during the last two years of his life. He began putting on weight in 1805. This became very noticeable in 1815; he eventually became obese and died so. This important statement, particularly emphasized by the advocates of the poisoning theory, requires consideration since it is the focus of a well-known contradiction. Several members of Napoleon’s entourage noted his loss of weight, not just Montholon – who according to Ben Weider mentioned it in order to avert the suspicion of poisoning. Whilst noting that Napoleon was “extraordinarily corpulent” before his death, Antommarchi also said that “he (Napoleon) has lost a lot of weight”. On another occasion, he insisted even more on Napoleon’s weight loss: “He has lost a lot of weight, he is emaciated”. According to him, Napoleon’s limbs and chest were wasted but not so much his stomach. Arnott also found that the patient’s legs and thighs were thin. On 16 April, 1821, Arnott found Napoleon “thinner than the last time he had had the occasion to see him”. At the autopsy, Antommarchi maintained that “the Emperor had lost a lot of weight since his arrival at St Helena (and that) he was less than a quarter of the size he was before”. Furthermore, it is known that some measurements of Napoleon’s body were taken at the autopsy by Antommarchi himself. One day, Napoleon said to Antommarchi, talking about his legs: “(…) they are worn out (…); you see, nothing is left: it is just bones”. To which Antommarchi replied that such fragility was a consequence of the disease. In his Memoirs, Marchand gives a similar testimony. Montholon himself, talking about Napoleon’s illness: “He is as thin as in 1800 and I look big and fat compared to him”. Thus, no doubt remains, Napoleon did lose weight.
One point, however, has not been adequately emphasized. Napoleon did not die of advanced cancer, he died of a complication of cancer, in his case a gastrointestinal hemorrhage. At the autopsy, no metastases were found except in regional lymph nodes related to the stomach. Metastases usually accompany cancer in the advanced stage. Whilst this latter detail would imply that the cancer was probably no longer limited only to the stomach itself, nevertheless it was in the early stage of spread, and it was therefore too early to expect dramatic weight loss in the patient. Weight loss always accompanies stomach cancer in its advanced stage. Such is also the case in chronic arsenic poisoning, contrary to René Maury (who claims that “obesity is equally a distinctive sign of chronic arsenic poisoning”) and Ben Weider who produces as evidence a compilation of symptoms of arsenic poisoning, listed without references to the sources. I have consulted a number of works on this subject. They all state that considerable weight loss is symptomatic of chronic arsenic poisoning. At the time of Napoleon’s death, due to the limited evolution of his cancer, emaciation was not yet manifest, although evident weight loss had been observed by Montholon and others, shortly before his death and at the autopsy.
Let us now consider the question of the regional lymph nodes of the stomach noted at the autopsy and mentioned above. Forshufvud claims that corrosion of the stomach precipitated by a strong dose of calomel can present all the appearances of a cancer. Corrosion of the stomach possibly, but certainly not a hardening and enlargement of the regional lymph nodes. And yet, such was the finding at the autopsy as described by Antommarchi, who was, remember, the only pathologist present. Hard and enlarged lymph nodes strongly indicate metastatic spread of the cancer, which means the first stage of dissemination. At a later stage, the dissemination affects other organs, in particular the liver, but Napoleon’s premature demise due to an internal hemorrhage precluded this development.
The only absolute proof that Napoleon did have stomach cancer would be a microscopical analysis of the relevant tissue. This would also be possible on mummified tissue. Nowadays, this is the only way to confirm cancerous growth. Unfortunately, we know that Napoleon’s descendants refuse any access to his remains.
I would like to put forward some other arguments against arsenic poisoning.
It is true that the weakness of the legs affecting Napoleon could be considered as a manifestation of neuropathy, that is, a nerve disorder generated by arsenic, but this condition may accompany other diseases – such as cancer – and only a clinical examination would allow an unequivocal diagnosis. It should be remembered in this context that Napoleon suffered many injuries – particularly to the legs – which pained him both during and after his military career.
According to General Bertrand, Napoleon in agony lost consciousness never to regain it. We remember his extreme weakness during the last days of his life, and the rapid pulse noted in agony. These symptoms have been interpreted in favour of arsenic poisoning. However, an internal hemorrhage produces the same effects, as well as a state of shock. Napoleon also suffered from lasting and painful nausea. Both often accompany stomach cancer. This argument has also been used, wrongly in my opinion, in favour of the poisoning hypothesis.
I have mentioned several times the arguments advanced by René Maury in favour of poisoning. In order to corroborate them, he refers, amongst others, to the very famous toxicologist Professor Léon Derobert and quotes passages from his book published in 1984, Intoxications and professional diseases. Professor Derobert points out that, as widely known, arsenic has a carcinogenic effect. However, certain preconditions need to be present: a long term exposure and in low doses, as well as a predisposition to cancer, hereditary or acquired. And he adds: “The cancer always develops on the basis of pre-existent hyperkeratosis”. René Maury refers to this statement in order to support his argumentation on the role of arsenic in the development of stomach cancer. Yet, pre-existent hyperkeratosis only relates to epidermoid cancers, mainly of the skin, and not of the stomach.
Furthermore, the autopsy did not reveal (apart from for the stomach – which was severely ruined) any of the internal pathology associated with arsenic ingestion and it is therefore not possible to confirm the poisoning theory. It is well known that chronic arsenic poisoning generates malignant tumors, above all of the liver. Yet, Napoleon’s liver was healthy.
The chronic depression Napoleon suffered from, seems to me to be another element which may be advanced in favour of cancer. Napoleon was predisposed to depression. Drained by melancholy, overwhelmed by sadness, regrets and the indifference of posterity, he always carried with him a dose of poison in case he should succumb to the temptation to commit suicide; something which, in fact, he attempted in 1814. The moral anxiety created by deliberate humiliations, petty slights and the stress aggravated by melancholy and forced inactivity could have created – according to recent medical beliefs – could have created a situation favourable the development of cancer. Yet, this detail has not been addressed by any of the scholars.
It would be out of place and above all pointless to repeat the other arguments advanced by some experts supporting the cancer hypothesis. I simply refer those interested to the many publications on this subject. The ambition of this study is to contribute a different element to the discussion, which might, I hope, settle it by proving that Napoleon died of complications of stomach cancer precipitated by the administration of a deadly dose of calomel.
It is appropriate, in this context, to draw the attention to the considerable number of expressions such as, amongst others, probably, perhaps, this could mean that …, one cannot rule out that …, there is a good chance that …, there is a chance of probability, it is not historically certain, these elements in fact point towards the possibility of …, etc. etc., found in the writings of those who opt for poisoning. Ben Weider and René Maury seem to be the only ones never have doubts or queries which would allow interpretations differing from theirs. Yet, science as we all know, is a complex and often delicate exercise. As regards my theories concerning Napoleon’s cancer, there remain, here too, some clouded patches, notably the impossibility of an unequivocal interpretation of the symptoms of his illness from 1817 on.
The duration of symptoms – four years – has been seen as favoring poisoning, and was also regarded by the doctors present at the autopsy as a factor in support of hepatitis. It is true that a period of four years seems too long for a stomach cancer, which usually produces symptoms only for a year or so before its final stage. But on the other hand, we must not forget that it was only from September 1820 – that is, eight months before his death – that interrelated symptoms appeared, pointing to the development of one single disease. Stomach cancer is often preceded by chronic gastritis or by a stomach ulcer, and the symptoms caused by these conditions can last a long time, even for years. Since a stomach ulcer was indeed found at Napoleon’s autopsy, the duration of his illness does not necessarily have to be interpreted as long term poisoning as Forshufvud and Weider have it, but could be considered as the result of a chronic gastritis leading to cancer.
Finally, I think it important to emphasize that in the field of toxicology considerable progress has been made over the last 80 years in methods of detection of arsenic, but the results still remain difficult to interpret, as demonstrated, amongst others, by the Besnard case, in the fifties.
I also confess that during my work on this paper, I have many times felt in danger when confronting certain arguments differing from my own, in particular those proposed by Forshufvud. I equally admit that I have occasionally been seduced by the logic of his reasoning and misled at the same time by the ambiguity of certain vague and contradictory symptoms and of some incidents which occurred among the Emperor’s entourage. In this light, it is not possible for me to discredit totally the arsenic poisoning line, but I am convinced that it is a matter of coincidence and that Napoleon did indeed suffer from cancer and that he died prematurely from acute complications thereof, and directly by the administration of a deadly dose of calomel.
It is a commonplace in medicine that every interpretation of symptoms of whatever disease is liable to be subjective, both on the part of the patient and on that of the physician. This is why only laboratory tests such as radiology and biochemistry could lead – and even then not infallibly! – to an unequivocal diagnosis. Given the importance and status of Napoleon, there was always going to be the risk that medical considerations would be modified for political reasons. This is why only an updated review of the autopsy, the three independent reports of which converge, seemed to us likely to allow us to come closer to the truth because, regrettably, the histopathological techniques in this case are presently beyond our reach.