49

The first question that arises with the measurement of the Earth’s circumference by the Greek Eratosthenes (3rd c. BC), is the length of the stade he used. Based on little known works, the author shows that the 158 m proposed in 1802 by Girard, a member of the French expedition to Egypt, and confirmed in 1972 by the analysis of Strabo’s Geography conducted by the Russian Firsov, should be preferred over all other values, and particularly, the 185 m suggested by Martin in 1854. With a stade of 158 m, the 252,000 stades of Eratosthenes’ measurement are less than 0,5%from today’s known value. This remarkable accuracy, which has given rise to abundant literature, has led to belief in a precision of the same order. But accuracy is not precision. The four rough approximations that Eratosthenes was obliged to perform, given the rudimentary technical means then at his disposal and the method applied – two on distances and two on angles – almost compensated one another perfectly. Consequently, the author shows that the precision (or rather the imprecision !) of Eratosthenes’ measurement was, on the contrary, rather of 20 to 25%.