BLOGS

The predictions were amazingly precise. Each day was divided into three or more segments, each of which was given a rating lying somewhere in the range from very favourable to highly adverse.

One of the best preserved of these papyrus documents is called the Cairo Calendar. Although the papyrus is badly damaged in places, scholars have been able to extract a complete list of ratings for days throughout an entire year somewhere around 1200 BC.

An interesting question is how the scribes arrived at their ratings. So various groups have studied the patterns that crop up in the predictions. Today, Lauri Jetsu and buddies at the University of Helsinki in Finland reveal the results of their detailed statistical analysis of the Cairo Calendar. Their conclusion is extraordinary.

These guys arranged the data as a time series and crunched it with various statistical tools designed to reveal cycles within it. They found two significant periodicities. The first is 29.6 days--that's almost exactly the length of a lunar month, which modern astronomers put at 29.53059 days.

The second cycle is 2.85 days and this is much harder to explain. However, Jetsu and co make a convincing argument that this corresponds to the variability of Algol, a star visible to the naked eye in the constellation of Perseus.

Algol is interesting because every 2.867 days, it dims visibly for a few hours and then brightens up. This was first discovered John Goodricke in 1783, who used naked eye observations to measure the variability.

Astronomers later explained this variability by assuming that Algol is a binary star system. It dims when the dimmer star passes in front of the brighter one.

Nothing else in the visible night sky comes close to having a similar period so it's reasonable to think that the 2.85 and the 2.867 day periods must refer to the same object. "Everything indicated that the two best periods in [the data] were the real periods of the Moon and Algol," say Jetsu and co.