Part of the Archimedes book series (ARIM, volume 4)


China has the longest astronomical heritage of any country of the world. Systematic astronomical records began to be kept in the eighth century BC and continue more or less uninterrupted up to the present day. Whilst astronomy in Babylon, the only other great civilization in the ancient world from which a vast array of astronomical records is preserved, had more or less ceased by the first century AD, the traditional astronomy of China that had developed by the Han dynasty (c. 200 BC), continued up until the start of the present century. That is not to say, however, that astronomy in China was not influenced at times by other cultures. For example, during the Yuan dynasty, many Islamic astronomers were invited to the Chinese capital, and an Islamic Observatory was set up.1 More significantly, when the Jesuits came to China in the seventeenth century AD they brought with them western astronomical knowledge with which they helped to reform the Chinese calendar.2 However, throughout all of this, Chinese astronomy retained a character all of its own. Furthermore, it was to be at the root of all of the astronomy which was to develop in the two other great ancient and medieval cultures of East Asia, Japan and Korea.


Solar Eclipse Ming Dynasty Chinese History Lunar Eclipse Small Mark 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Yabuuti & van Dalen (1997).Google Scholar
  2. 2.
    For discussions of the activities of the Jesuit astronomers in China, see d’ Elia (1960) and Bernard (1973).Google Scholar
  3. 3.
    Ho (1970).Google Scholar
  4. 4.
    Nakayama (1969: 18–19).Google Scholar
  5. 5.
    Needham (1954).Google Scholar
  6. 6.
    Needham (1954: 134).Google Scholar
  7. 7.
    Xu, Yau & Stephenson (1989); Xu, Stephenson & Jiang (1995).Google Scholar
  8. 8.
    Stephenson & Yau (1992).Google Scholar
  9. 9.
    Han (1955).Google Scholar
  10. 10.
    In modem Chinese usage, T’ien-wen means astronomy, but in the context of early Chinese history, it may be more appropriate to translate the term as astrology. See Nakayama (1966: 442). This Table is based upon the work of Han (1955).Google Scholar
  11. 12.
    Yabuuti (1973).Google Scholar
  12. 13.
    Yabuuti (1974).Google Scholar
  13. 14.
    Cullen (1993).Google Scholar
  14. 15.
    The gaps in the Table relate to dates where it is not clear which calendar was adopted.Google Scholar
  15. 16.
    Yabuuti (1979).Google Scholar
  16. 17.
    An English translation of the Hui-hui-li with commentary is currently being prepared by B. van Dalen.Google Scholar
  17. 18.
    The apparent precision of this value simply comes from a minor alteration of the value from the preceding Ssu fenli of 365 math = 365 5math . See Sivin (1969: 12). 19 A list of the values of the constants in the various calendars is given by Yabuuti (1963a, 1963b).Google Scholar
  18. 20.
    Yabuuti (1963a: 460–470).Google Scholar
  19. 21.
    Liu & Stephenson (1998).Google Scholar
  20. 22.
    Yabuuti (1963a: 462).Google Scholar
  21. 23.
    Yabuuti (1963a: 469).Google Scholar
  22. 24.
    Sivin (1986).Google Scholar
  23. 25.
    For example, in Liu Hsiang’s Wu-ching T’ung-i. See Needham (1959: 414).Google Scholar
  24. 26.
    This is one of several arguments made against the “astronomical” explanation of eclipses by Wang Ch’ung in chapter 32 of his Lun Heng. See Forke (1907: 269) and Needham (1959: 411–413).Google Scholar
  25. 27.
    Translations of the relevant parts of these calendars are given by Sivin (1969). In addition, he gives a detailed discussion of the two (astronomically equivalent) methods described in the Ssu -fen-li with worked examples. For additional discussions, see Eberhard & Müller (1936) and, briefly, Aaboe (1972).Google Scholar
  26. 28.
    Rounding down in this way implies that the initial eclipse was at one extreme of the nodal zone. However, this implication is not explicitly stated in the text, and may not have been considered by the early Chinese astronomers.Google Scholar
  27. 29.
    In practice, of course, this would break down since the relation 23 eclipse possibilities = 135 months is not without error.Google Scholar
  28. 30.
    For a general discussion of these cycles, see Sivin (1966, 1969).Google Scholar
  29. 31.
    Yabuuti (1973).Google Scholar
  30. 32.
    Yabuuti (1963a: 476).Google Scholar
  31. 33.
    Yabuuti (1963a: 479).Google Scholar
  32. 34.
    The k’o is a unit of time such that there are 100 k’o in one day. See Section 6.5 below.Google Scholar
  33. 35.
    Nakayama (1969: 144–145).Google Scholar
  34. 36.
    Yabuuti (1963a: 482–483).Google Scholar
  35. 37.
    See Chiu T’ang-shu, 35 for a description of the work. This has been translated and discussed by Beer et al. (1961), who remark that it is “one of the most remarkable pieces of organized field research in the ear ly middle ages: ‘Google Scholar
  36. 38.
    Sivin (1997) has translated the Shou-shih-li I, a discussion of the calendar contained in the Yuan-shih, into English in an unpublished manuscript. A detailed discussion of the Shou-shih-li is given by Nakayama (1969: 123–150).Google Scholar
  37. 39.
    Or, alternatively, 4 math marks in each of the initial and central halves of a double hour if this system was being used.Google Scholar
  38. 40.
    Maspero (1939).Google Scholar
  39. 41.
    Maspero (1939).Google Scholar
  40. 42.
    It should be noted that Maspero analyses the sunrise and sunset timings to try to determine their accuracy, having already assumed that the small mark was at the end of the double hour on the basis of the limited written evidence discussed above.Google Scholar
  41. 43.
    Although, of course, the distribution of the errors is certainly not random. At least in part, this is probably a factor of the interpolation used in calculating the length of day and night throughout the year. This and the following graphs seem to suggest that the Chinese customarily used a ratio for the longest to the shortest night which was slightly too small.Google Scholar
  42. 44.
    This is explicitly stated in the second set of times and may be inferred for the first set from the times themselves which would otherwise have sunrise and sunset an hour earlier than the true value.Google Scholar
  43. 45.
    Needham, Lu, Combridge, & Major (1986: 9).Google Scholar
  44. 46.
    Needham (1959: 315–317).Google Scholar
  45. 47.
    Needham (1959: 327).Google Scholar
  46. 48.
    Forte (1988).Google Scholar
  47. 49.
    The Hsin I Hsiang Fa Yuo has been translated into English by Needham, Wang & de Solla Price (1986).Google Scholar
  48. 50.
    Needham (1965: 494–498).Google Scholar
  49. 51.
    Bo (1997).Google Scholar
  50. 52.
    Needham 1965: 508).Google Scholar
  51. 53.
    Hua (1997).Google Scholar
  52. 54.
    See Stephenson (1994) for a list of the lunar lodges.Google Scholar
  53. 55.
    Some commentators, for example Stephenson (1997a), have interpreted phrases such as Chih wei hou 1 k’o as meaning “the 1 st mark in the central half of the hour of wei.” However, the practice of splitting the double hour into an initial and a central half did not come into general use until later times and a more likely reading is “after the 1 st mark of the hour of wei,” as I have given above. I am grateful to Dr. Liu Ciyuan of Shaanxi Observatory, China, for a helpful discussion of this issue.Google Scholar
  54. 56.
    I follow Sivin (1997) in the translation of these terms.Google Scholar
  55. 57.
    Or, at least, are not among the limited collection of work that is readily available to modem scholars.Google Scholar
  56. 58.
    Mo-k’o Hui-hsu, 7; trans. Needham, Wang, & de Solla Price (1996: 16).Google Scholar
  57. 59.
    We might speculate that the records in the Yuan-shih, which were compiled by an author involved in calendar reform, are more likely to have come from the Bureau of Astronomy and Calendar where he presumably worked, albeit many years later, and the records in the Wen-hsien T’ung-k’ao to have come from the Astronomical Department of the Imperial Academy. However, there is no firm evidence either way.Google Scholar
  58. 60.
    Despite P’eng Ch’eng’s comment that the astronomers in the observatories “contented themselves with copying out the positions of the sun, moon, and planets according to very rough ephemerides ... never using the (astronomical equipment in the) observatory” [Mo-k’o Hua-hsi, 7; trans. Needham, Wang & de Solla Price (1986: 16)], these very inaccurate times cannot have been calculated. The Sung calendars were able to predict the time of an eclipse to at least within an hour. Furthermore, calculating an eclipse time using these methods is not an easy task — certainly requiring more effort than actually observing the eclipse.Google Scholar
  59. 61.
    Stephenson & Fatoohi (1995); Fatoohi & Stephenson (1996). .Google Scholar
  60. 62.
    And, for the lunar eclipses, to the greater length of the night watches in winter than in summer.Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2000

Authors and Affiliations

  1. 1.University of DurhamUK

Personalised recommendations