Published: Journal of Analytical Toxicology, ISSN 0146-4760, Volume 26, Number 8, November/December, pp. 584-585

LETTER TO THE EDITOR: A New Series of Hair Analyses from Napoleon Confirms Chronic Exposure to Arsenic
Pascal Kintz1, Jean-Pierre Goullé2, Paul Fornes3, and Bertrand Ludes1
1Institut de Médecine Légale, 11 rue Humann, 67000 Strasbourg, France; 2Laboratoire de Pharmacologie et de Toxicologie, CHG, Le Havre, France; and 3Médecine Légale, Faculté de Médecine Cochin-Port Royal, Paris, France

To the Editor:
Napoleon Bonaparte died on the island of Sainte Hélène on May 5, 1821.
Since the end of the 20th century, there has been a debate on the manner of death of the Emperor, as revealed by numerous publications, including several books. The official thesis is that Napoleon died from carcinoma of the stomach complicated by gastric bleeding that may have been precipitated by the large dosage of calomel given the day before he died.
Since 1964 and the presentation (1) of the results of the analysis (using neutronic activation) of Napoleon’s hair by the Harwell Nuclear Research Laboratory of the University of Glasgow that demonstrated high arsenic content, there has been a great controversy about the final conclusions involving a criminal poisoning. In 1999, these results were reviewed by Weider and Fournier (2), who confirmed the arsenic poisoning issue.
Several hairs attributed to the Emperor have been tested around the world, and in most cases, they exhibited arsenic content in higher amounts than the physiological concentrations of the present time. The normal arsenic background for the early 19th century is not known. In 1998, Hindmarsh and Corso (3) published a compendium of these results (about 20 analyses).
The arsenic poisoning thesis was contested by the official version, according to the following elements: 1. the hair specimens did not belong to Napoleon; 2. an external contamination (including contamination through the wallpaper, exposure to coal smoke, the use of arsenic-containing cosmetics, or the use of arsenic to preserve hair samples) can explain positive arsenic results; and 3. arsenic ingestion (e.g., arsenic-containing tonic or arsenic occuring naturally in the water supply) is possible (4–6).
In this particular climate of controversy among people actively involved in history and biology, in December 2002 our laboratory received five hair specimens attributed to Napoleon. These specimens were presented by Dr. Ben Weider, President of the International Napoleonic Society (Montréal, QC, Canada), with the request to test for arsenic.
The following specimens were submitted to the Laboratory of Toxicology of the Institute of Legal Medicine of Strasbourg (France):
Hair identified as “Las Cases”, cut by Santini on October 16, 1816, and retained by Las Cases. Some of these hairs were given to Mr. William Fraser (Dehli, India) in an envelope. This envelope belonged to Mr. Troubetskoy (USA), who gave it to
Dr. Hamilton Smith. We tested two hairs from this series.
Hair identified as “Abbé Vignali”, cut on May 6, 1821, and stored by the priest.
Hair identified as “lady Holland”, cut on May 6, 1821, and stored in a golden box that was given to lady Holland. The strand was fixed on the center of a medallion.
Hair identified as “Abram Noverraz”, cut on May 6, 1821. These specimens belonged to the museum of Arenenberg (Switzerland), which permitted us nine hairs. These specimens are the longest ever tested, some measuring 9 cm.
Hair identified as “Louis Marchand”, cut on May 6, 1821, and belonging to Mr. Jean Tranié, who gave us three hairs.
Hair specimens were thoroughly decontaminated according to the following procedure: acetone (5 mL, 2 min), warm water (5 mL, 2 min), and finally acetone (5 mL, 2 min) with horizontal shaking.
After weighing, hair samples (0.5 to 2.2 mg) were hydrolyzed in 60 µL of 1M NaOH at 90°C for 30 min. Neutralization was achieved with the addition of a small amount of acid (60 µL of 1.03M nitric acid). Before testing by atomic absorption, 400 µL of matrix modifier (nickel nitrate at 6 g/L of nickel) was added.
Total arsenic was measured by graphite furnace atomic absorption spectroscopy with reading at 193.7 nm (window at 0.5 nm) using a Varian SpectrAA-800 system equipped with Zeeman correction.
Using a 2-mg sample, the response of the apparatus was linear from 0.5 to 120 ng arsenic per milligram of hair. In these conditions, the limit of detection of the system was 0.11 ng/mg. Intraday precision, evaluated with 20 consecutive analyses, was 9.7%.
Results of the toxicological investigations are reported in Table I, ranging between 6.99 and 38.53 ng/mg hair.
Because of the small amount of biological material that was submitted to us, it was not possible to use ICP-MS
(this technique requires at least a 100-mg sample) to test for arsenic or other elements, nor was it scientifically acceptable to cut the hair into small segments to establish the pattern of arsenic exposure. It seems difficult to test for arsenic with classical atomic absorption spectroscopy using some micrograms of hair, in order not to measure the background noise. Moreover, the idea to use the calculation of a hair growth of 0.8 to 1.3 cm/month to establish a timeline of exposure seems unreliable because of irregular growth, drug movement in the hair shaft, and difficulties in this particular case to determine the orientation (i.e., the root and the tip).
A compendium (7) of numerous recently published papers indicates that the physiological concentrations of arsenic in hair are less than 1.0 ng/mg (average 0.5 ng/mg), whereas concentrations in subjects with chronic poisoning are often in the 1–5-ng/mg range. Based on this statement, it seems obvious that the hair specimens attributed to Napoleon demonstrate chronic exposure to arsenic.
An extensive decontamination procedure was used. This is of paramount importance to avoid external deposition of arsenic along the hair shaft.
According to the biography of the Emperor, he did not use any cosmetic treatment containing arsenic. A postmortem contamination (e.g., from the earth of the cemetery, such as in the Marie Besnard case) can also be excluded because the hair specimens were collected on the day Napoleon died or while he was alive.
To demonstrate the efficacy of the decontamination procedure, a control strand of hair obtained from a co-worker in the laboratory was incubated for 4 h in a As2O3 aqueous solution at 10 mg/L. Before contamination, arsenic concentration was
0.31 ng/mg. After contamination and without decontamination, the arsenic concentration was 9.86 ng/mg, which was eliminated after the triple stage (acetone/warm water/acetone) decontamination, with a resulting concentration of 0.34 ng/mg.
The positive response to arsenic could be also explained by the following: use of pharmaceuticals containing arsenic (however, no historian reported such a practice), drinking water contaminated with naturally occuring arsenic in Longwood [this water was analyzed by Hindmarsh and Corso (3), and arsenic concentration was < 0.002 ppm], contamination from wallpaper pigments containing arsenic (the wallpaper was changed in the drawing room in 1819, and therefore cannot be blamed for the arsenic measured in the strand of hair identified as “Las Cases”, cut in 1816), and finally, contamination by exposure to coal smoke (only Napoleon developed clinical symptoms, the other people living in this area did not). As all these parameters can be discussed and eliminated, it seems acceptable that arsenic poisoning occured by oral administration.
Differences in arsenic concentrations from the four specimens cut on May 6, 1821, can be explained by differences in length of the hair, differences in growing stages (only hair in the anagen stage can incorporate xenobiotics, in comparison with the catagen or telogen stages) of the hair and finally differences in anatomic sites of collection from the head.
Using graphite furnace atomic absorption spectroscopy, five hair specimens attributed to Napoleon exhibited arsenic concentrations substantially higher than physiological concentrations. These concentrations are conclusive for arsenic poisoning. A strong decontamination procedure allowed us to avoid problems relative to external contamination.
In order to definitely validate these findings, one should be able to attribute the analyzed hair to Napoleon through DNA tests and to measure arsenic content in the hair of some relatives who were in Longwood during the same period.

Reproduction of editorial content of this journal is prohibited without publisher’s permission.

| Home | Subscribe | Current Issue | Back Issues | Search | Advertise | Other Publications |