N.S. Gill at About.com has a nice article on the
Etymology of Geometry Terms where she points to the rather troublesome etymology of our word CIRCLE, writing:
"That circle (coming either from a Greek verb meaning to hoop around or from the circular Roman circus) is marked with what you would have, in pre-geometry days, called a line across part of it."
That purported etymology is VERY WEAK and so is the linguists theory that the Indo-European root of circle is sker-, for which there is as good as no evidence.
Actually, I think the origin of the word CIRCLE comes from an even more ancient Indo-European word, that for "millstone", found, for example, in Latvian word DZIRKALIS "millstone", which did not get its original name from its shape at all, but rather from the function it served.
In Latvian DZIR- would have the same origin as Latvian DZIRKSTELLE "spark" and KALIS can mean "Smith, Smithy" in Latvian relating to the root KAL-T "to forge, to hammer" and so DZIRKALIS, the "millstone" would have some origin such as "spark grinder", a term probably going far back into the history of man and the observations that certain stones ground together would produce sparks - i.e. fire. Presumably, millstones were the first circular objects practically used and it is likely, in our opinion, that these ultimately led to the idea of the wheel for other applications (spinning wheels, etc.).
In time, the millstone became synonymous with the concept of a circle and we thus later obtained many words for circular concepts with the root CIRC-, which does not have an original concept of circle as its base, but rather a millstone for grinding.
