In 1849 cholera killed 14,007 people in London's 38 registration districts. William Farr blamed the low ground. John Snow blamed the water. I rebuilt both arguments from their own figures, measured every death on Snow's Broad Street map, and asked whether the worst-hit districts are still the poorest today.
Cholera reached England and Wales in 1848, and over the next two years about 52,000 people there died of it. It killed unevenly. In Rotherhithe, on the low south bank of the Thames, 205 people in every 10,000 died. In Hampstead, on its hill, it was 8.
William Farr, the statistician at the General Register Office, set out the figures district by district in his 1852 report and found a beautifully regular pattern: the higher the ground, the fewer the deaths. He read it as the work of bad air rising from the riverside soil. John Snow, a doctor, read the same years differently: cholera was something people swallowed, and the low districts were the ones drinking the Thames.
This page puts their numbers side by side. Three questions run through it. Does Farr's elevation law hold up? Was it really about the water? And does the map of 1849 still show in the map of London's incomes today? Each death in these tables was a person; I have tried to keep that in view.
Each shape is one of Farr's districts (a few small neighbours are merged), coloured by deaths per 10,000 people in 1849. Switch to today's income to see the same ground in 2023. The shapes are my approximation from named places, not historic boundaries: see the methods.
Farr's table gives each district's height above the Thames high-water mark. The pattern is as strong as he said. Across the 38 districts the rank correlation between height and the death rate is -0.81. The 16 districts below 20 feet lost 91.7 people per 10,000; the two at 100 feet or more lost 16.3, 5.6 times fewer.
Farr fitted a law to it: a smooth curve on which mortality falls as the ground rises, from 174 per 10,000 at the river to 6 at 350 feet. A curve of the same shape fitted to his own districts explains 68% of the variation in death rates, weighting each district by its population.
The trouble with Farr's law is that height and water went together. Farr's own table sorts the districts into three water sources. The twelve districts drinking Thames water taken between Battersea and Waterloo Bridge, in central London, were all low: none stood higher than 22 feet. They lost 114.4 people per 10,000. Districts supplied from the Thames upstream at Kew and Hammersmith lost 17.8, 6.4 times fewer.
Farr's table on its own cannot pull the two apart. Height alone accounts for 63% of the variation in (logged) death rates across districts; water source alone for 58%; the two together for 75%. Height still matters once water is allowed for, but less: each doubling of height cut deaths by 30% on its own and by 26% after water. Water sources were defined coarsely, and low districts were poorer and more crowded too, so this is a tangle, not a verdict.
The verdict came from south London. There, two companies, the Southwark and Vauxhall and the Lambeth, served the same districts house by house, so that neighbours could drink different water. By 1854 the Lambeth company was drawing its water from further up the Thames, at Seething Wells, while the Southwark and Vauxhall still took it from a stretch fouled by London's sewage. The ground had not moved. The water had.
In 1849 the two companies' customers died at the same rate, 12.2 per 1,000. In 1854 Southwark and Vauxhall customers died at 13.5 per 1,000 and Lambeth customers at 3.6: 3.8 times fewer. Snow's own count for the first seven weeks of 1854 was starker still: 315 deaths per 10,000 houses on Southwark and Vauxhall water, 37 on Lambeth water.
In 1849 both companies' customers died at the same rate. By 1854 one company had moved its intake upriver, and its customers' death rate had fallen by 70%.
Five years later, at the end of August 1854, cholera broke out around Broad Street in Soho. Snow marked each death as a small bar at its address. This map shows the 578 deaths on his map, placed by the georeferenced copy in the cholera R package, and the 13 public pumps.
Half of the deaths were within 117 metres of the Broad Street pump as the crow flies, and 88% within 200 metres. Walking along the streets, 375 of the 578 deaths (65%) were at addresses closer to Broad Street than to any other pump. Within 100 metres of the pump the dead lay at 71.6 per hectare; between 200 and 300 metres, 3.8.
On my measure the other 203 deaths were nearer, on foot, to another pump, many of them only just. Snow, who walked these streets himself, counted only 10 deaths in houses "decidedly nearer" another pump. Of those, 5 families told him they always sent to Broad Street because they preferred its water, and 3 were children who went to school near it. The map cannot show that. It was the conversations that made the case.
The handle came off on 8 September, after Snow spoke to the parish guardians. By then the outbreak had already turned: 315 fatal attacks began between 31 August and 2 September, and only 28 on 7 September. Many people had fled. Snow himself wrote that the attacks had fallen so far before the handle came off that it was impossible to decide whether the well was still dangerous. What it shows is how short and how local the disaster was.
To compare 1849 with today I gave each of London's 2021 small areas (MSOAs) to the 1849 district whose named places lie nearest, on the same bank of the river. That gives 31 areas, covering 353 MSOAs and 2,981,956 people in 2021, each with its 1849 death rate and its average household income in 2023.
There is no link. The rank correlation between the 1849 death rate and income after housing costs today is -0.07, which is nothing. The eight worst-hit areas of 1849 now average £46,700 a year after housing costs; the eight least hit, £41,200. Wandsworth, where 99.9 in 10,000 died, now has the second highest income after housing costs of the 31 areas; Paddington, which lost 8.5, ranks 21st.
But the 1849 map of wealth has left a faint mark. Farr also recorded the value of houses per person, a rough measure of how well off a district was. That does line up with income today: the rank correlation with total household income in 2023 is 0.46 (p = 0.01 across 31 areas). Bethnal Green had the cheapest housing per head of all 31 areas in 1849 and has the second lowest income after housing costs now. Cholera has left London's map; the geography of who could afford what has not.
London was only part of it. Farr also printed the daily deaths from cholera for all of England and Wales in 1849: 53,293 in the year. The worst day was 6 September, with 1,121 deaths; August and September alone held 68% of the year's total. I looked for an open digital copy of Farr's district table for the whole of England and Wales, so as to map the national picture, and did not find one, so this study stays in London.