On Antimony in Pipes and Pompeii: Yeah … About That

Yesterday evening came the first word of a study bringing up the old idea of ‘lead and the downfall of Rome’ with a different twist. Specifically, that it wasn’t lead that was a big deal at Pompeii, but rather antimony, which can have all sorts of nasty ill effects in various contexts.  It is rather problematic as I see it, and hopefully this post will prevent the sensationalism which is already starting to percolate around this study. Let’s begin with the press release from the University of Southern Denmark:

The ancient Romans were famous for their advanced water supply. But the drinking water in the pipelines was probably poisoned on a scale that may have led to daily problems with vomiting, diarrhoea, and liver and kidney damage. This is the finding of analyses of water pipe from Pompeii.

 The concentrations were high and were definitely problematic for the ancient Romans. Their drinking water must have been decidedly hazardous to health.

This is what a chemist from University of Southern Denmark reveals: Kaare Lund Rasmussen, a specialist in archaeological chemistry. He analysed a piece of water pipe from Pompeii, and the result surprised both him and his fellow scientists. The pipes contained high levels of the toxic chemical element, antimony.

The result has been published in the journal, Toxicology Letters.

Romans poisoned themselves

For many years, archaeologists have believed that the Romans’ water pipes were problematic when it came to public health. After all, they were made of lead: a heavy metal that accumulates in the body and eventually shows up as damage to the nervous system and organs. Lead is also very harmful to children. So there has been a long-lived thesis that the Romans poisoned themselves to a point of ruin through their drinking water.
However, this thesis is not always tenable. A lead pipe gets calcified rather quickly, thereby preventing the lead from getting into the drinking water. In other words, there were only short periods when the drinking water was poisoned by lead: for example, when the pipes were laid or when they were repaired: assuming, of course, that there was lime in the water, which there usually was, says Kaare Lund Rasmussen.

Instead, he believes that the Romans’ drinking water may have been poisoned by the chemical element, antimony, which was found mixed with the lead.

Advanced equipment at SDU

Unlike lead, antimony is acutely toxic. In other words, you react quickly after drinking poisoned water. The element is particularly irritating to the bowels, and the reactions are excessive vomiting and diarrhoea that can lead to dehydration. In severe cases it can also affect the liver and kidneys and, in the worst-case scenario, can cause cardiac arrest.

This new knowledge of alarmingly high concentrations of antimony comes from a piece of water pipe found in Pompeii.

– Or, more precisely, a small metal fragment of 40 mg, which I obtained from my French colleague, Professor Philippe Charlier of the Max Fourestier Hospital, who asked if I would attempt to analyse it. The fact is that we have some particularly advanced equipment at SDU, which enables us to detect chemical elements in a sample and, ever more importantly, to measure where they occur in large concentrations.

Volcano made it even worse

Kaare Lund Rasmussen underlines that he only analysed this one little fragment of water pipe from Pompeii. It will take several analyses before we can get a more precise picture of the extent, to which Roman public health was affected.

But there is no question that the drinking water in Pompeii contained alarming concentrations of antimony, and that the concentration was even higher than in other parts of the Roman Empire, because Pompeii was located in the vicinity of the volcano, Mount Vesuvius. Antimony also occurs naturally in groundwater near volcanoes.

This is what the researchers did

The measurements were conducted on a Bruker 820 Inductively Coupled Plasma Mass Spectrometer.

The sample was dissolved in concentrated nitric acid. 2 mL of the dissolved sample was transferred to a loop and injected as an aerosol in a stream of argon gas which was heated to 6000 degrees C by the plasma.

All the elements in the sample were ionized and transferred as an ion beam into the mass spectrometer. By comparing the measurements against measurements on a known standard the concentration of each element is determined.

The story was picked up practically verbatim by the usual science sites:

It’s also been picked up elsewhere and rewritten:

So much for coverage. If one goes to the actual article and downloads it (for 36 bucks), there will be profound disappointment, as the press release is probably as long as the article (fortunately I have a son with access). Here’s the citation if you care to pursue it (it’s considered ‘in press, corrected copy’):

P. Charlier, F. Bou Abdallah, R. Bruneau, S. Jacqueline, A. Augias, R. Bianucci, A. Perciaccante, D. Lippi, O. Appenzeller, K.L. Rasmussen, Did the Romans die of antimony poisoning? The case of a Pompeii water pipe (79 CE), Toxicology Letters, 2017, ISSN 0378-4274, http://dx.doi.org/10.1016/j.toxlet.2017.07.876.

As is usually the case, I will preface my objections/observations by pointing out I’m a Classicist, not a scientist, but there are a couple of things in this study which set off alarm bells: one from a Classics standpoint, and one from a Science standpoint. First, the Classics objection: the piece of pipe that was studied was “a small metal fragment of 40 mg, which I obtained from my French colleague, Professor Philippe Charlier of the Max Fourestier Hospital.” As I voiced briefly on Twitter a couple of times, this is not really acceptable, even as a starting point. How does a Professor in a French hospital come to have a piece of Pompeii pipe? The Toxicology Letters article says it comes from the House of Caecilius Jucundus from 1875 and is now in a private collection. I really think we need some documentation confirming that provenance. Yes, the House of Caecilius Jucundus was excavating in 1875, but how did this tiny piece of pipe come to be in France? There are plenty of sources of Roman lead to be found, and our current knowledge of the antiquities market suggests we need more assurance in regards to how this piece got to France from Pompeii, if indeed it did.

The other objection is tied more to the piece in Toxicology Letters. Without getting into lead and calcification of pipes which has recently been called into question (see, e.g., D. Keenan-Jones, Lead contamination in the drinking water of Pompeii) we note this comment in the article, dealing with the ‘leachability’ of antimony into a water supply from lead pipes:

This is an alarming level because antimony is easily leached from the lead water pipe, and antimony poses a serious health hazard, as demonstrated by a study conducted on loops of lead pipes containing twenty Sn/Sb joints soldered by antimony. Barely detectable initially (less than 0.004 ppm), antimony concentration in the water running through the pipes reached a level of 0.01 ppm in 4 days, and of 0.068 ppm within four weeks (corresponding to 0.068 mg/L) (Murrell, 1987). This effect has not been reproduced by other studies due to the variability of experimental conditions such as water pH, debit, and concentration of antimony in the pipes.

The first thing to note is that the comparative study is based on pipes which were soldered with antimony, so the water flowing through those pipes presumably would come in direct contact. Roman pipes were soldered with lead, not antimony. Second, we seem to be dealing with some sort of closed system involving ‘loops’, so a particular quantity of water would be repeatedly exposed to the antimony solder as opposed to just flowing through a pipe, presumably picking up more antimony each time around. That looks like a situation which will skew results. Finally, and perhaps most importantly, the effect is apparently not reproducible and hasn’t been reproduced in thirty years. If you’re interested in tracking down that study (which Google couldn’t find today):

Murrell, N.E., 1987. Impact of Metal Solders on Water Quality. Proceedings of the Annual Conference of the American Water Works Association, Part 1. Denver American Water Works Association pp. 39–43.

That said, it’s worth noting that the author of the study is aware of the limitations based on one sample, but that admission will be forgotten in the sensational headlines. In the press release:

Kaare Lund Rasmussen underlines that he only analysed this one little fragment of water pipe from Pompeii. It will take several analyses before we can get a more precise picture of the extent, to which Roman public health was affected.

The piece in Toxicology Letters concludes with:

We strongly suggest further studies (including antimony level analyses in human bones and coprolites) to verify this groundbreaking theory.

In other words, by all accounts we’re still at the theory stage. This really shouldn’t be getting press coverage yet, much less sensational headlines.