It is uncertain whether the well at Syene, in Upper Egypt, which he used for this purpose, was dug by his directions, or existed previously.  Pliny seems to be of the former opinion; but there is reason to believe that it had a much higher antiquity.  The following observations on its structure by Dr. Horsley, Bishop of Rochester, are ingenious and important.  “The well, besides that it was sunk perpendicularly, with the greatest accuracy, was, I suppose, in shape an exact cylinder.  Its breadth must have been moderate, so that a person, standing upon the brink, might safely stoop enough over it to bring his eye into the axis of the cylinder, where it would be perpendicularly over the centre of the circular surface of the water.  The water must have stood at a moderate, height below the mouth of the well, far enough below the mouth to be sheltered from the action of the wind, that its surface might be perfectly smooth and motionless; and not so low, but that the whole of its circular surface might be distinctly seen by the observer on the brink.  A well formed in this manner would afford, as I apprehend, the most certain observation of the sun’s appulse to the zenith, that could be made with the naked eye; for when the sun’s centre was upon the zenith, his disc would be seen by reflection on the water, in the very middle of the well,—­that is, as a circle perfectly concentric with the circle of the water; and, I believe, there is nothing of which the naked eye can judge with so much precision as the concentricity of two circles, provided the circles be neither very nearly equal, nor the inner circle very small in proportion to the outer.”

Eratosthenes observed, that at the time of the summer solstice this well was completely illuminated by the sun, and hence he inferred that the sun was, at that time, in the zenith of this place.  His next object was to ascertain the altitude of the sun, at the same solstice, and on the very same day, at Alexandria.  This he effected by a very simple contrivance:  he employed a concave hemisphere, with a vertical style, equal to the radius of concavity; and by means of this he ascertained that the arch, intercepted between the bottom of the style and the extreme point of its shadow, was 7 deg. 12’.  This, of course, indicated the distance of the sun from the zenith of Alexandria.  But 7 deg. 12’ is equal to the fiftieth part of a great circle; and this, therefore, was the measure of the celestial arc contained between the zeniths of Syene and Alexandria.  The measured distance between these cities being 5000 stadia, it followed, that 5000 X 50 = 250,000, was, according to the observations of Eratosthenes, the extent of the whole circumference of the earth.