Abstract

This paper investigates the methods of eclipse prediction in China before the fourth century AD, with a detailed example of the eclipse theory in the Jing chu li (Luminous Inception System ). As the official calendar of the Jin dynasty and the Kingdom Wei during the three kingdoms period, the Jing chu li was used for more than 200 years after it was adopted in 237 AD. From the San tong li (Triple Concordance System ) of the Western Han to the Jing chu li, methods for predicting eclipses developed in three important ways: (i) from predicting only lunar eclipses to the prediction of both solar and lunar eclipses; (ii) from relying only on the mean periods of the sun and the moon to taking into consideration the variation in lunar velocity; and (iii) from estimating only a rough date to predicting the exact time of eclipses. This paper addresses two questions: first, how did ancient Chinese astronomers use cycles to predict eclipses in the Han dynasty? Second, how did astronomers such as Liu Hong and Yang Wei revise early eclipse prediction methods? The original text of the Jing chu li is analyzed to show how Yang Wei combined lunar velocity theory with the traditional method of predicting eclipses using cycles.