Scientists have never been able to pin down when Egypt's remarkable Old Kingdom pyramids were built. Nor could they figure out how a few were so exquisitely aligned along a north-south axis without the benefit of modern surveying equipment.
Now a British researcher has managed to answer both these questions. She believes she can date the start of construction for each pyramid to within five years, and reveals how the ancients pulled off a most exceptional feat in construction engineering.
The answers are all in the stars, according to Dr Kate Spence of the University of Cambridge, who describes her research this morning in the science journal, Nature. A simple piece of string and two stars from the Big and Little Dipper constellations (Ursa Major and Ursa Minor, to purists) gave the Egyptians a method that could be repeated for every pyramid. And with a bit of computation her finding also tells us when each pyramid was started.
All the pyramids of the time are aligned on a north-south axis, Dr Spence points out, but the accuracy varies wildly from about 1/20th of a degree to a full degree off line. The most accurate is the Pyramid of Khufu (Cheops) at Giza. So close is it to true north that scientists have marvelled at the achievement since it was discovered in the 19th century.
Researchers had looked at stellar and solar alignment methods but none fully explained what was going on. Why was it that Cheops could be so accurate while the others drifted wide of the mark? Why couldn't the Egyptians replicate their feat?
Dr Spence argues that the answer lies in a concept called "precession", a term which describes how the earth "wobbles" over time as it orbits the sun. The planet behaves like a huge toy gyroscope that spins but can't maintain a rigid orientation. It tips off the vertical in a 26,000-year cycle which can be calculated.
The pyramids' builders, she argues, used a pair of fairly bright stars - Kochab in Ursa Minor and Mizar in Ursa Major. In the year 2,467 BC these stars pointed due north when the two formed a vertical line. All the Egyptians had to do was use a weight and a string and wait for the two stars to line up as the night progressed. When both disappeared behind the string the viewer was looking due north.
Construction on the Great Pyramid at Cheops therefore began within five years either side of 2,467 BC, she believes. This is a great improvement on previous best estimates, which have a margin of error of 100 years either way.
Her work also neatly explains why pyramids built before and after this date seemed to drift off the north-south axis. It is all down to that orbital wobble. "The most likely explanation of this deviation from north is that a stellar orientation method was used which became increasingly inaccurate as a result of the effects of precession," she writes in Nature.
Those built earlier than Cheops are inaccurate to one side of true north, while those built afterwards are inaccurate on the other side of true north. Because the error in the alignment can be readily calculated for any date, Dr Spence's method allows her to date each pyramid very finely.
Such accuracy is of tremendous value to Egyptologists, who study the pyramids and the kings who occupied them. It allows the specialists to get a better understanding of when and how long each pharaoh reigned and a more accurate picture of the chronologies for the Old Kingdom period of 5,000 years ago.
Dr Spence believes she can do even better. Her plan is to refine her work to achieve a margin of error down to within one or two years at most.