From that data, Eratosthenes was able to draw a map of the world on a grid of parallels and meridians, basing his longitudes on a meridian running through Alexandria and his latitude on a line passing through the Pillars of Hercules and Rhodes.
The result would probably have been a reasonably accurate map of the Mediterranean, but the map has been lost, as have the three volumes of his Geography although fragments of the project can be pieced together from the works of writers like Marcianus, Pliny, Polybius and Strabo.
In his version of things, the inhabited earth, the oikoumene, ended at the Ganges, which was around one-third of the way around the globe on the Rhodes parallel. Beyond the Ganges, and beyond the Pillars of Hercules at the western end of the parallel, there was Ocean. Eratosthenes did not deny the possibility of land beyond the extremities of the oikoumene, or in the southern temperate zone. Since their existence was a matter of speculation and supposition, there was no way he could include them on his map, but he seems to have assumed that the temperate zone south of the equator would have been inhabited like its northern equivalent.
Eratosthenes also commented on the nature and origin of the Earth, which he thought was an immovable globe with a changing surface. He hypothesised that at one time the Mediterranean was a vast lake that had only become connected to the ocean to the west when a passage had opened up at the Pillars of Hercules.
His conclusions, however, were not universally accepted:
Eratosthenes goes into a description of the figure of the earth; not merely of the habitable earth, an account of which would have been very suitable, but of the whole earth, which should certainly have been given too, but not in this disorderly manner. He proceeds to tell us that the earth is spheroidal, not however perfectly so, inasmuch as it has certain irregularities, he then enlarges on the successive changes of its form, occasioned by water, fire, earthquakes, eruptions, and the like; all of which is entirely out of place, for the spheroidal form of the whole earth is the result of the system of the universe, and the phenomena which he mentions do not in the least change its general form; such little matters being entirely lost in the great mass of the earth. Still they cause various peculiarities in different parts of our globe, and result from a variety of causes.