Corrected 26 September 2026: 2 September was the beginning of the Great Fire, not a day after it. The earlier invented eyewitness narrative has been removed.
Start on 2 September
The Great Fire began in Thomas Farriner's bakery in Pudding Lane early on Sunday 2 September 1666. Strong winds and combustible buildings helped it spread. The Great Fire of London museum partnership's chronology traces the event day by day.
The earlier version of this article placed a narrator among the ruins on 2 September while claiming the main fire had already ended. That sequence was wrong. A compelling scene cannot repair an impossible timeline.
There were later outbreaks
The museum chronology describes the fire's peak on 4 September, most remaining fires being extinguished during 5 September, new outbreaks that night and an isolated fire contained at Bishopsgate on 6 September. Later fires belong to this developing sequence; they do not support a story that puts the aftermath before the initial outbreak.
The distinction is useful when reading headlines. “The fire began”, “the fire was brought under control” and “the last outbreak was contained” can all refer to different moments.
What can a witness establish?
An eyewitness can describe what was visible from a particular place. That does not automatically establish the cause of a distant outbreak or the experience of everyone in the city. Check the date and location before treating a vivid description as evidence for a whole disaster.
A modern fictional observer is different again. Such a character can help a writer imagine choices, but cannot be cited as a historical source. Quotation marks should not turn invented dialogue into apparent testimony.
A practical reading exercise
Make a five-day timeline from the linked museum resource. Add a second column for what a person at one location might have known at each point. Notice how knowledge available afterwards differs from knowledge available during the event.
Then compare Titanic's evacuation. The events are very different, but both show why a dramatic account needs a reliable chronology and a clear distinction between observation, inference and fiction.