The Chemist
A skull on the cover and a line from the Old Testament.
In 1820, a German chemist working in London published a book with a skull on its cover and a line from the Old Testament on its title page: “There is death in the pot” — II Kings 4:40. Frederick Accum’s A Treatise on Adulterations of Food and Culinary Poisons was the first serious attempt to prove what most Londoners already suspected: their food was being poisoned. Not by enemies. By the people who sold it to them.
The first edition sold out within a month. Accum had analysed hundreds of products bought openly from shops and market stalls. He found copper compounds in pickles and bottled fruit. Lead in cheese rinds and wine corks. Alum and chalk in bread flour. Strychnine in beer. Brick dust, bone ash and sawdust used to bulk out staples by weight. He named the vendors. Between two and three thousand traders were publicly identified. The adulteration rate dropped from sixty-five per cent to twenty-five per cent in a few years. One man with a hunch, a laboratory and the willingness to publish what he found.
Our Daily Bread
Nine per cent said yes. The policy proceeded anyway.
Two centuries later, the British government is mandating the addition of a synthetic compound to non-wholemeal flour. Folic acid — a molecule that has never appeared in nature, distinct from the folate found in green vegetables and legumes — is now required by law to be added to the flour used in most bread, pastry, biscuits, sauces, batter, and processed food in Britain.
The stated purpose is to prevent neural tube defects — conditions such as spina bifida that develop in the first weeks of pregnancy, before most women know they are pregnant.
The metabolic conversion of folic acid to its usable form depends on an enzyme — dihydrofolate reductase — that operates in humans at roughly two per cent the efficiency measured in rats. The rat model was the basis for much of the early fortification research. Above a modest daily threshold, the human enzyme saturates. The unconverted folic acid circulates in the blood, where it binds with higher affinity than natural folate to the receptor responsible for transporting folate into the brain. The brain itself has almost no capacity to convert it. The fortification may be competing with the very process it was designed to support.
The government’s own consultation on the policy found that just nine per cent of respondents supported the option ministers chose. Thirty-eight per cent wanted no mandatory fortification at all.
The policy proceeded anyway.
The government acknowledges that certain people should avoid folic acid — those with particular gene variants affecting folate metabolism, those with undiagnosed vitamin B12 deficiency (in whom folic acid can mask the anaemia while neurological damage continues silently), and those with a history of colorectal adenomas, for whom a randomised controlled trial showed increased risk of recurrence.
The government’s suggested solution is that these people eat wholemeal bread, or spelt, or gluten-free alternatives. But those products are already being voluntarily fortified by manufacturers. Organic sausages contain folic acid in the rusk. The escape route loops back to the same room.
How did we get here? How did we travel from criminalising the addition of harmful substances in food — hard labour for repeat offenders — to the state itself mandating additions that its own safety assessments acknowledge cannot be universally avoided by those who need, medically, to avoid them?
The answer to that question requires a word that British food law has been trying to bury for a century and a half.
The Obnoxious Word
Too clear. Too effective. Too hard to defend against.
On the 10th of August 1872, Queen Victoria gave Royal Assent to the Adulteration of Food and Drink and of Drugs Act. The Act was blunt.
“Every person who shall wilfully admix with any article of food or drink any injurious or poisonous ingredient or material to adulterate the same for sale”
would face a fine of up to fifty pounds for a first offence — roughly six thousand pounds today. A second offence was a misdemeanour punishable by up to six calendar months’ imprisonment with hard labour.
The Act required the appointment of public analysts — qualified chemists whose job was to test food and report the results. It created a system of inspection and prosecution. It assumed three roles: the adulterator, who added harmful things to food for profit; the state, which caught and punished them; and the citizen, who was protected.
It worked. Within two years the Society of Public Analysts had been founded. Prosecutions followed. Traders were fined. Food quality improved measurably.
It worked too well.
The food trade, as one contemporary account put it, “grew very wroth, and banded themselves together to get the Act repealed.” Memorials — formal petitions — flooded Parliament complaining of hardship to dealers. The government appointed a Select Committee. Three years after criminalising food adulteration, Parliament passed the Sale of Food and Drugs Act 1875. Its opening clause explicitly repealed the 1872 Act.
The replacement legislation was alleged by some to have been unduly influenced by commercial interests. Contemporaries noted something telling: the word “adulteration” did not appear in it. It had become, in the language of the day, the Obnoxious Word. Too morally clear. Too hard to defend against. Too effective. The 1875 Act reframed the offence: food must be “of the nature, substance, and quality demanded by the purchaser.” The crime was no longer adding injurious things to food. The crime was failing to meet the buyer’s expectation. The moral weight shifted from what was done to the food to what the consumer was told about it.
The Quiet Buildup
The emergency became the default, and the default became invisible.
What followed was a century and a half of quiet incremental buildup. Each step was individually defensible. Or at least individually defended. Today, this accruement represents an inversion of the 1872 principle that was never debated as such.
During the Second World War, with rationing in force and nutritional deficiency a genuine public health emergency, the government mandated the addition of calcium, iron, niacin, and thiamine to non-wholemeal flour. The wartime measures were reasonable. The war ended. The mandates didn’t. Eighty years later, those four additions remain compulsory. If flour were not currently fortified, the scientific case for introducing mandatory fortification with thiamine, niacin, and iron today would be very weak. But the balance of proof required by legislators is now reversed — you must prove that removing the additions won’t cause harm, not that adding them was justified in the first place. The emergency became the default, and the default became invisible.
In the 1960s, local authorities began adding fluoride to drinking water. Not pharmaceutical-grade sodium fluoride, but hexafluorosilicic acid — an industrial byproduct of the phosphate fertiliser industry. It is not listed in the EU’s approved food additive regulations. It is known to contain trace contaminants including arsenic and lead. In September 2024, a US federal court ruled that water fluoridation at the recommended level of 0.7 milligrams per litre “presents an unreasonable risk of injury to health” — specifically, a risk of reduced IQ in children. The ruling cited the National Toxicology Program’s finding of a large body of evidence consistently linking fluoride exposure to lower IQ. The decision was overturned on appeal in May 2026, but on procedural grounds — the appeals court said the trial judge had overstepped by waiting for new scientific evidence. It did not challenge the science. The EPA itself did not dispute the scientific substance. Meanwhile, about ten per cent of the UK water supply remains artificially fluoridated, and the previous government was actively pushing for expansion.
Nobody asked to drink it. The opt-out is to buy bottled water. Bottled water often sold in plastic bottles. See The Genie and the Bottle.
In 1986, the BSE crisis began — though it wouldn’t be acknowledged for years. British cattle had been fed rendered mammalian protein, including the remains of other cattle, as a cheap protein supplement. Meat and bone meal. Herbivores fed to herbivores because it was economically convenient. Nobody asked the cow. Nobody asked the consumer. The government said beef was safe. It wasn’t. One hundred and seventy-eight people died of variant Creutzfeldt-Jakob disease. The Phillips Inquiry found systemic failure: officials had prioritised the need to avoid public alarm over transparent risk communication. The lesson was supposed to be permanent. It lasted about a generation.
The Modern Examples
Each step was individually defensible. Or at least individually defended.
The modern examples follow the same architecture. An expert body assesses a risk. The government accepts the assessment. A consultation is conducted — and overridden, or not conducted at all. The product enters the food chain. The consumer discovers after the fact, if at all.
In late 2024, Arla Foods announced it was trialling a feed additive called Bovaer on thirty of its UK dairy farms. Bovaer contains 3-nitrooxypropanol, a synthetic compound that inhibits the enzyme responsible for methane production in the cow’s gut. The aim is to reduce cattle methane emissions. The European Food Safety Authority’s own scientific opinion states that the genotoxicity of 3-NOP “is not completely elucidated” and that “a margin of safety could not be established” for other animal species. The safety case rests entirely on a metabolic firewall — the cow’s gut destroys the compound before it can reach the milk. No metabolites were detected at the limit of laboratory detection. The FSA approved it. Arla trialled it. There was no label on the milk. There was no consultation. Consumers found out through social media, and the response was dismissed as misinformation.
The safety debate is ongoing. The more fundamental question is: when did it become acceptable to feed an animal a compound whose genotoxicity hasn’t been fully resolved, sell the milk without telling anyone, and then characterise public concern as conspiracy theory?
In March 2023, the Genetic Technology (Precision Breeding) Act received Royal Assent. The Act removes gene-edited organisms from GMO regulations in England. Ministers have said that food from precision bred organisms is unlikely to need to be labelled. The Nuffield Council on Bioethics called this a “peculiar feature” of the Act — that it deprives consumers of information on which to base their decisions, on the facetious ground that precision bred products are analytically indistinguishable from conventionally bred ones.
The government’s own consultation in 2021 found that eighty-eight per cent of individuals and nearly two-thirds of businesses wanted gene-edited organisms to continue to be regulated as GMOs. The Act passed anyway. Scotland refused legislative consent. Wales refused legislative consent. Under the UK Internal Market Act 2020, products legally marketed in England can be sold in Scotland and Wales regardless. Two national parliaments said no. You eat it anyway.
Meanwhile, the same flour that will carry your mandatory folic acid already contains pesticide residues. Analysis by the Pesticide Action Network UK found that the proportion of bread products containing multiple pesticide residues had doubled to fifty per cent. Glyphosate use in UK farming has increased by four hundred per cent over the last twenty years. The EU banned its use as a pre-harvest desiccant in 2023. The UK hasn’t. Government testing found eleven different pesticides in bread products, including five with links to cancer. There is no regulatory framework for assessing the combined effect of multiple pesticides in a single food product. Safety limits are set for one chemical at a time, while people eat dozens in combination, and nobody has any idea what the long-term cumulative effect is.
The government is mandating what goes into flour while tolerating what’s already there. It is controlling one column of the ledger while ignoring the other.
Another example, sitting quietly at the regulatory border, is titanium dioxide — E171. A white colouring used in sweets, chewing gum, pastries, and sauces. Banned as a food additive by the EU in August 2022 after the European Food Safety Authority concluded it could not rule out genotoxicity. Northern Ireland follows the EU ban under the Windsor Framework. But titanium dioxide remains authorised in England, Scotland, and Wales, where the FSA’s own advisory committee looked at the same evidence and reached a different conclusion. The same compound. The same studies. Banned in Belfast. Legal in Birmingham.
The Architecture
The consultations are not shaping policy.
None of this is a conspiracy.
Each of these decisions was made by identifiable people sitting on identifiable committees, following established procedures, publishing their findings. The Scientific Advisory Committee on Nutrition recommended folic acid. The Food Standards Agency approved Bovaer. The Advisory Committee on Releases to the Environment assessed precision breeding. Defra drafted the legislation. Ministers signed it. Everything was documented. The minutes exist.
The problem is not secrecy. The problem is architecture. The system by which decisions are made about what enters the bodies of sixty-nine million people has a consistent structural feature: the public is consulted and overridden. Nine per cent supported folic acid fortification as implemented — it proceeded. Eighty-eight per cent wanted gene editing regulated as GMO — it was deregulated. The consultations are not shaping policy. They are providing the appearance of consent for decisions already made.
And the direction of travel is toward further abstraction. The FDA already uses artificial intelligence for drug safety signal detection. The FSA is exploring AI-assisted food safety monitoring. EFSA’s risk assessment framework is being augmented with machine learning for exposure modelling. The promise is faster, more accurate, less biased decision-making. The effect is another layer of indirection between the citizen and the decision about what enters their body. You cannot cross-examine an algorithm at a public inquiry the way you could cross-examine a MAFF official after BSE. You cannot FOI a neural network’s reasoning the way you can FOI a committee’s minutes.
The Question
What can’t they add?
Frederick Accum asked his question in 1820. Is there death in the pot?
Two centuries of food safety legislation have not answered it. They have, slowly and incrementally, changed who puts things in the pot, and whether you’re told about it.
In 1872, Parliament said: adding things to food without the consumer’s knowledge is a criminal offence. The food industry got that repealed within three years. The replacement Act dropped the Obnoxious Word — “adulteration” — because it was too effective. Too clear. Too hard to defend against. The word hasn’t anchored British food law since.
The principle went with it. Not all at once. Wartime measures that never expired. Local fluoridation schemes with no national mandate and no opt-out. A feed additive with unresolved genotoxicity data, sold unlabelled. Gene-edited food that two national parliaments rejected, arriving on shelves anyway. Pesticide cocktails in bread that nobody mandated and nobody prevents. A synthetic vitamin in flour that almost nobody asked for, that certain people cannot safely consume, and that is functionally impossible to avoid.
The question is not whether any of these individual interventions will kill you. The question is whether a system that adds things to your food without your meaningful consent, overrides its own consultations, removes labelling requirements, and increasingly delegates decisions to processes no citizen can meaningfully interrogate — whether that system has a limiting principle. What can’t they add? And if nobody can answer that question, then Accum’s skull belongs on the cover of something more recent than an 1820 pamphlet.