Snow on the Broad Street pump
📚SOURCE
💎PERMANENT NOTES
🧵SYNTHESIS
Public domain (1855), so three of the highlights are quoted exactly. The rest is my summary of 🔬 On the Mode of Communication of Cholera.
Snow believed cholera was swallowed, not breathed. Most of his profession blamed bad air. The book is his case, and it has two famous parts of very unequal weight.
The Broad Street outbreak of 1854 is the story everyone knows: hundreds dead within a short walk of one pump, and a handle removed. It is vivid and it is local. One pump, one neighbourhood.
The stronger evidence is what he calls his experiment on the grandest scale. Two water companies served the same south London streets, pipes side by side, customers mixed house by house. One had moved its intake upriver, above the city's sewage. The other had not. Snow counted cholera deaths by supplier. The households on the dirty supply died at more than eight times the rate.
The map tells you where. The comparison tells you why.
✨KEY IDEAS
“The most terrible outbreak of cholera which ever occurred in this kingdom, is probably that which took place in Broad Street, Golden Square, and the adjoining streets, a few weeks ago.”
“In consequence of what I said, the handle of the pump was removed on the following day.”
Of the people in the two water companies' districts he writes that they “were divided into two groups without their choice, and, in most cases, without their knowledge”.
In the first seven weeks of the 1854 epidemic he counts 315 deaths per 10,000 houses supplied by the Southwark and Vauxhall company, 37 per 10,000 for the Lambeth company, and 59 for the rest of London. (My paraphrase of his table.)
The exceptions support him: the brewery workers who drank beer and the workhouse with its own well were largely spared, and a widow in Hampstead who had Broad Street water sent to her died. (My paraphrase.)
📝READING NOTES
How I came to it
Adaeze said in class that everyone knows the pump and nobody has read the book, and that the book is better. She was right on both counts. I read it yesterday on the train and finished it this morning in a hospital corridor while they changed Mum's dressing, and wrote this at her kitchen table afterwards.
Broad Street, as he tells it
- Late August 1854. Deaths cluster around one street pump in Soho.
- He gets the list of the dead and walks to the addresses. Nearly all lived closer to that pump than to any other, or were known to have fetched water from it.
- He takes this to the parish board on the evening of 7 September. The handle comes off on the 8th.
- He does not claim it ended the outbreak. He says attacks had already diminished, and that so many residents had left that he cannot tell what removing the handle achieved. The legend has him stopping an epidemic. The author says he may have prevented a second one. See 💡 The Broad Street pump handle was removed in September 1854.
The famous map came later, drawn to illustrate the argument and added to this second edition. He did not find the pump by looking at a map. He found it by knocking on doors.
The two companies
This is the half Adaeze meant.
| Houses | Deaths | Per 10,000 houses | |
|---|---|---|---|
| Southwark and Vauxhall | 40,046 | 1,263 | 315 |
| Lambeth | 26,107 | 98 | 37 |
| Rest of London | 1,422,570 | 1,422 | 59 |
What makes it evidence and not just a table:
- The two supplies were intermingled, often along the same street, so neighbourhood, income and air cancel out.
- Nobody chose their supplier with cholera in mind. Many did not know which one they had.
- Where the household did not know, he tested the water himself: the downstream supply was salty enough to tell apart chemically.
A natural experiment, on about three hundred thousand people, by one doctor on foot.
What I think the lesson is
The map is persuasive because it is a picture, and it proves only that the deaths were near the pump. A miasma would predict the same picture. The comparison is what a miasma cannot explain. I have been showing people maps of where exceptions cluster and wondering why nobody changes anything.
Where I would be careful: Snow was right, and it is tempting to read the book as if rightness were obvious. He had no germ, no mechanism he could show, and his contemporaries mostly remained unconvinced for years. Good evidence was not enough to persuade them. That is the less comfortable half of the claim I am about to write.
What it produced
💎 Evidence persuades when it compares, with this and Nightingale's diagrams as the two examples. The first half of the thought is in 📖 Spiegelhalter on asking a question of data.
More on the man in 🏛️ John Snow.
Added 2 December: presented the course project tonight. Since Hana's catch in November I think of Snow's households whenever two groups are compared: did they choose which group to be in.
Related notes
- 🏛️ John Snow
- 💡 The Broad Street pump handle was removed in September 1854
- Adaeze Nwosu
- 📖 Spiegelhalter on asking a question of data
Reading record
- Recommended by: Adaeze Nwosu, in class
🗣️IN SHORT
- In : On the Mode of Communication of Cholera, second edition (1855), the Broad Street section and the comparison of the south London water companies
- Snow argues that : cholera is carried in water contaminated by the sick, not in the air
- by stating that : deaths in Soho centred on one pump with telling exceptions, and above all that households on a sewage-tainted supply died at many times the rate of their neighbours on a clean one, though the two groups differed in nothing else