rotation of Mercury
Antoniadi's map of Mercury, c. 1920
The planet Mercury is notoriously difficult to observe, being both small and always in the glare of the Sun as seen from Earth. At best it can be studied telescopically for a few weeks each year, near to sunrise or sunset. However, undeterred by these problems, astronomers from the late eighteenth century on did their best to glimpse features on Mercury's surface and, using these as fixed reference points, obtain the length of the Mercurian day.
From his observations of the innermost planet between 1779 and 1813, Johann Schröter deduced the existence of a mountain 20 km high and a Mercurian rotation period of just over 24 hours. Later, William Denning reported surface markings "so pronounced that they suggest an analogy with those of Mars" and adjusted the period to 25 hours. But it was Giovanni Schiaparelli's suggestion, that Mercury's day might be 88 Earth days long, the same as its year, which was generally accepted from the 1880s until as recently as the early 1960s. Concomitant with this hypothesis of gravitational lock was the intriguing idea of a (possibly habitable) "twilight zone" (see Mercury, life on). Based on the assumption that Schiaparelli's period was correct, Eugene Antoniadi produced a map of dark and light areas of the planet, complete with wonderfully evocative names, such as Aurora, Apollonia, Pieria, Liguria, and Cyllene. The first sign of trouble for the Schiaparellian model came in 1962, when measurements by W. E. Howard and his colleagues at Michigan revealed that the night side of Mercury was warmer than it should have been if it never faced the Sun. Three years later, radar measurements, by Gordon Pettengill and Rolf Dyce of Cornell University using the Arecibo radio telescope, showed conclusively that Mercury is not locked in position relative to the Sun. It is in a gravitational resonance, but one in which its axial rotation period is about 58.6 Earth days, or two thirds of its orbital rotation period.