The Bedford Level experiment: what was tested and what was seen
Check it yourself · Updated
The Bedford Level experiment is a series of observations along a straight, 6-mile stretch of the Old Bedford River in England, used to test whether a water surface curves. Samuel Rowbotham reported in 1838 that a boat stayed visible down to the waterline. Later tests on the same canal reported curvature, and refraction explains why results differ.
Rowbotham’s experiment, 1838
The Old Bedford River in Cambridgeshire is a canal over 20 miles long and almost perfectly straight, with still water and no locks. Rowbotham had a boat with a flag 3 feet above the water sail from Welney Bridge to Welche’s Dam, 6 miles apart. He watched through a telescope while sitting in the water, with his eye no more than 8 inches above the surface, and reported that “the flag and the boat down to the water’s edge were clearly visible throughout the whole distance”. He calculated that on a globe the boat would have been more than 16 feet below his line of sight.
The numbers
| Setup (6 miles, eye 8 in) | Hidden below the horizon |
|---|---|
| Globe, no refraction | about 16.7 ft (5.1 m) |
| Globe, standard refraction (k = 0.13) | about 14 ft (4.3 m) |
| Flat surface | 0 ft |
Try these and your own numbers in the Earth curvature calculator.
Later tests on the same canal
- 1870, Alfred Russel Wallace. In a public wager with John Hampden, Wallace set two markers at the same height, about 13 feet above the water, on bridges 6 miles apart, with a third marker halfway. Through a telescope the middle marker appeared higher than the other two, as expected on a curved surface. The referee gave the wager to Wallace; Hampden disputed the result for years.
- 1901, Henry Yule Oldham. Repeated the three-marker test and reported the same raised middle marker.
- 1904, Elizabeth Blount. With a photographer, reported that a white sheet at the far end could be seen and photographed from near the water, a result like Rowbotham’s.
Why the results differ: refraction
Air is usually denser near the ground, which bends light slightly downwards and lets you see a little beyond the geometric horizon. Over cool water on a calm day the effect can be much stronger and can even lift objects that would otherwise be hidden. Observations made inches above the surface, like Rowbotham’s, are the most affected. That is why later tests raised the line of sight to about 13 feet and compared three markers instead of looking at one distant object.
How to try it yourself
Use a long, straight stretch of water and a target of known height. Look from as low and as high as you can, on a cool morning and a warm afternoon, and compare what you see with the Earth curvature calculator. Write down the time and the weather: they change the result.
Frequently asked questions
What did the Bedford Level experiment show?
Rowbotham (1838) reported that a boat 6 miles away stayed fully visible, which he took as proof of flat water. Alfred Russel Wallace (1870) and Henry Yule Oldham (1901) reported that a middle marker appeared raised, as on a curved surface. Elizabeth Blount (1904) reported a result like Rowbotham's.
How much should be hidden over 6 miles?
With an eye 8 inches above the water, a globe would hide about 16.7 feet of an object 6 miles away without refraction, and about 14 feet with standard refraction. On a flat surface nothing would be hidden.
Why did the experiments give different results?
Air just above water bends light, sometimes strongly. Observations made close to the surface are the most affected, which is why later tests raised the line of sight to about 13 feet.