A prime meridian is a meridian (a line of longitude) in a geographic coordinate system at which longitude is defined to be 0°. Together, a prime meridian and its anti-meridian (the 180th meridian in a 360°-system) form a great circle. This great circle divides a spheroid, e.g., Earth, into two hemispheres. If one uses directions of East and West from a defined prime meridian, then they can be called the Eastern Hemisphere and the Western Hemisphere. A prime meridian is ultimately arbitrary, unlike an equator, which is determined by the axis of rotation—and various conventions have been used or advocated in different regions and throughout history. The most widely used modern meridian is the IERS Reference Meridian. It is derived but deviates slightly from the Greenwich Meridian, which was selected as an international standard in 1884. Longitudes for the Earth and Moon are measured from their prime meridian at 0° to 180° east and to 180° west. For all other Solar System bodies, longitude is measured from 0° (their prime meridian) to 360°. West longitudes are used if the rotation of the body is direct, that is, it follows the right hand rule. East longitudes are used if the rotation is retrograde. The notion of longitude was developed by the Greek Eratosthenes (c. 276 BC – c. 195 BC) in Alexandria, and Hipparchus (c. 190 BC – c. 120 BC) in Rhodes, and applied to a large number of cities by the geographer Strabo (64/63 BC – c. 24 AD). But it was Ptolemy (c. AD 90 – c. AD 168) who first used a consistent meridian for a world map in his Geographia. Ptolemy used as his basis the 'Fortunate Isles', a group of islands in the Atlantic which are usually associated with the Canary Islands (13° to 18°W), although his maps correspond more closely to the Cape Verde islands (22° to 25° W). The main point is to be comfortably west of the western tip of Africa (17.5° W) as negative numbers were not yet in use. His prime meridian corresponds to 18° 40' west of Winchester (about 20°W) today. At that time the chief method of determining longitude was by using the reported times of lunar eclipses in different countries. Ptolemy's Geographia was first printed with maps at Bologna in 1477, and many early globes in the 16th century followed his lead. But there was still a hope that a 'natural' basis for a prime meridian existed. Christopher Columbus reported (1493) that the compass pointed due north somewhere in mid-Atlantic, and this fact was used in the important Treaty of Tordesillas of 1494 which settled the territorial dispute between Spain and Portugal over newly discovered lands. The Tordesillas line was eventually settled at 370 leagues west of Cape Verde. This is shown in Diogo Ribeiro's 1529 map. São Miguel Island (25.5°W) in the Azores was still used for the same reason as late as 1594 by Christopher Saxton, although by then it had been shown that the zero magnetic deviation line did not follow a line of longitude. In 1541, Mercator produced his famous 41 cm terrestrial globe and drew his prime meridian precisely through Fuerteventura (14°1'W) in the Canaries. His later maps used the Azores, following the magnetic hypothesis. But by the time that Ortelius produced the first modern atlas in 1570, other islands such as Cape Verde were coming into use. In his atlas longitudes were counted from 0° to 360°, not 180°W to 180°E as is usual today. This practice was followed by navigators well into the 18th century. In 1634, Cardinal Richelieu used the westernmost island of the Canaries, Ferro, 19° 55' west of Paris, as the choice of meridian. The geographer Delisle decided to round this off to 20°, so that it simply became the meridian of Paris disguised. In the early 18th century the battle was on to improve the determination of longitude at sea, leading to the development of the marine chronometer by John Harrison. But it was the development of accurate star charts, principally by the first British Astronomer Royal, John Flamsteed between 1680 and 1719 and disseminated by his successor Edmund Halley, that enabled navigators to use the lunar method of determining longitude more accurately using the octant developed by Thomas Godfrey and John Hadley. Between 1765 and 1811, Nevil Maskelyne published 49 issues of the Nautical Almanac based on the meridian of the Royal Observatory, Greenwich. 'Maskelyne's tables not only made the lunar method practicable, they also made the Greenwich meridian the universal reference point. Even the French translations of the Nautical Almanac retained Maskelyne's calculations from Greenwich—in spite of the fact that every other table in the Connaissance des Temps considered the Paris meridian as the prime.'