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


Until the end of the last century the little that was known of Mesopotamian astronomy came from scattered references to the “Chaldaeans,” portrayed as mystical astrologers in the works of Greek authors such as Strabo and Diodorus Siculus and the Old Testament,1 and from the small number of observations contained in Ptolemy’s Almagest which he describes as having been observed in Babylon. However, after the archaeological exploration of Assyria and Babylonia by Austin Henry Layard and others which began in the 1840s, original cuneiform records have come to light that have transformed this image.2 Not only were celestial events used for divination in Mesopotamia, they were also systematically observed and recorded in Babylon from around the middle of the eighth century BC,3 and by the fourth century BC mathematical schemes had been developed that allowed various astronomical phenomena such as the appearance and disappearance of the planets and the time-intervals between successive oppositions and conjunctions of the sun and moon to be calculated.4 Babylonian astronomy was to have a profound and long lasting legacy: Greek, Indian, Islamic, and even Medieval European astronomy all drew on its influence.5


Solar Eclipse Lunar Eclipse Eclipse Observation Astronomical Diary Contact Interval 
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.
    See Neugebauer (1975: 607–614).Google Scholar
  2. 2.
    For a general history of the excavations and the recovery of the cuneiform texts, see Larsen (1996) and Budge (1925). Details of the recovery of the astronomical and astrological tablets are given by Neugebauer (1957: 53–70).Google Scholar
  3. 3.
    There is no firm evidence for similar long-term systematic observational programmes in the other Mesopotamian cities. However, during the eighth and seventh centuries BC, reports of astronomical observations and their astrological interpretations were regularly sent to the Assyrian kings in Nineveh.Google Scholar
  4. 4.
    Aaboe (1974) has described this final type of astronomy as “scientific,” that is “a mathematical description of celestial phenomena capable of yielding numerical predictions that can be tested against observations.”Google Scholar
  5. 5.
    For a general overview of the legacy of Babylonian astronomy in other cultures, see Pingree (1998) and Neugebauer (1963). For the records of Babylonian astronomical observations in Ptolemy’s Almagest, see Chapter 3. On the transmission of Babylonian mathematical astronomy to Greece, see Toomer (1988) and Jones (1991, 1993), and to India, see Pingree (1973, 1987).Google Scholar
  6. 6.
    Or, at least, near-original texts, as a number were obviously copied in antiquity by the Babylonian scribes. However, they are still only, say, second or third generation copies. By contrast, the writings from ancient China were written on perishable materials, and so we only possess late copies of the works that have gone through the copying process, with all of its inherent problems of corruption to the text, many times.Google Scholar
  7. 7.
    For a survey of the Enūma Anu Enlil series, see Weidner (1944a, 1944b, 1956, 1969) and KochWestenholz (1995). So far, the tablets concerned with lunar eclipses have been edited by Rochberg-Halton (1987, 1988), parts of those dealing with the planetary omens by Reiner & Pingree (1975, 1981, 1998), and those containing solar omens by van Soldt (1995).Google Scholar
  8. 8.
    The Venus Tablet has been published by Reiner & Pingree (1975).Google Scholar
  9. 9.
    For example, Langdon & Fotheringham (1928), Weir (1972), Huber et al. (1982), and Gasche et al. (1998) .Google Scholar
  10. 10.
    For a discussion of empiricism in the omen texts, see Rochberg (1999a).Google Scholar
  11. 11.
    Lambert (1975), Horowitz (1998).Google Scholar
  12. 12.
    MUL.APIN has been edited by Hunger & Pingree (1989).Google Scholar
  13. 13.
    It should be noted that although works such as Enūma Anu Enlil and MUL.APIN were written well before this time, they are mainly preserved in copies made during this later period.Google Scholar
  14. 14.
    This is a revised version of his earlier edition, Parpola (1970). See also his commentary, Parpola (1983). In the following discussions, the Letters are denoted by their LABS number in Parpola (1993). For a table of concordances with museum numbers, see Parpola (1993: 408–412).Google Scholar
  15. 15.
    This replaces the earlier edition by Thompson (1900). In the following discussions, the Reports are denoted by their ARAK number in Hunger (1992). For a table of concordances with museum numbers, see Hunger (1992: 374–379).Google Scholar
  16. 16.
    Hunger (1992: xvi).Google Scholar
  17. 17.
    The dates proposed by Parpola (1983) for many of the texts can not be justified. See, for example, de Meis & Hunger (1998) and Brown (1999b).Google Scholar
  18. 18.
    Reade (1986). 19 Sachs (1948).Google Scholar
  19. 20.
    In the following discussions, I will denote texts published in Neugebauer (1955) by their ACT number, and texts listed in Sachs (1955) by their LBAT number. LBAT numbers preceded by one star refer to tablets listed in Sachs (1955), but for which copies have not been published. Double starred LBAT numbers refer to texts listed in Sachs (1955) which have been published elsewhere — predominantly in Epping (1889), Kugler (1900, 1907, 1909, 1912, 1913, 1914, 1924), and Kugler & Schaumberger (1935). Texts not catalogued in either LBAT or ACT will be denoted by their museum numbers. Where joins have been made to previously catalogued fragmentary tablets, these are denoted by plusses. For lists of concordances between ACT, LBAT and museum numbers, see Neugebauer (1955: 453–459) and Sachs (1955: xi—lvi).Google Scholar
  20. 21.
    The term NMAT was coined by Aaboe (1980). 22 See, for example, the translation of Toomer (1984: 116).Google Scholar
  21. 23.
    Brinkman (1968: 227).Google Scholar
  22. 24.
    Hallo (1988).Google Scholar
  23. 25.
    The distribution of years with at least one dated Diary fragment is shown graphically in Figure 2 of Sachs (1974). Although a small number of additional Diary fragments have subsequently been dated, these would not significantly alter this figure.Google Scholar
  24. 26.
    These definitions are based upon those of Sachs & Hunger (1988: 20).Google Scholar
  25. 27.
    The Normal Stars are a group of bright stars located close to the ecliptic that were used as reference points in the sky. For a list of the Babylonian Normal Stars, see Sachs & Hunger (1988: 17–19).Google Scholar
  26. 28.
    Neugebauer (1947b, 1948). 29 Sachs (1952b).Google Scholar
  27. 30.
    Although Hunger (1999) has noted that there are a small number of cases where the observations recorded in the Goal-Year texts do not correspond exactly with the descriptions given in the Diaries, he has plausibly suggested that these Goal-Year texts drew on other Diaries than the ones that happen to be preserved, as there are cases where parallel Diaries give slightly different accounts of the same event.Google Scholar
  28. 31.
    See Sachs (1948: 283, 1955: xxv) and van der Waerden (1974: 108–110).Google Scholar
  29. 32.
    For BM 45728, see Kugler (1907: 45–48) and P. Huber in van der Waerden (1974: 107). For BM 41004, see Neugebauer & Sachs (1967: 200–208).Google Scholar
  30. 33.
    For this text, see Gadd (1967) and Hunger (1969).Google Scholar
  31. 34.
    However, Huber (1973) has edited most of the eclipse records found in the Goal-Year texts, together with many from the Diaries and the Eclipse Texts.Google Scholar
  32. 35.
    It should be noted that in many cases the records have obviously been reworded to give entries within a compilation uniform style.Google Scholar
  33. 36.
    Sachs (1976).Google Scholar
  34. 37.
    Sachs (1952a).Google Scholar
  35. 38.
    For references to additional texts, most of which have been published by Aaboe, Neugebauer, and Sachs, see Neugebauer (1975).Google Scholar
  36. 39.
    BM 40094, published by Aaboe (1969).Google Scholar
  37. 40.
    BM 34083 = ACT 53. 41 See, for example, Sachs & Hunger (1988: 25).Google Scholar
  38. 42.
    de Kuyper (1993).Google Scholar
  39. 43.
    Oppenheim (1969).Google Scholar
  40. 44.
    BOR 4, 132; CT 49, 144; and CT 49, 186. See, most recently, van der Spek (1985) and Rochberg (1999b). An earlier discussion is given by McEwan (1981: 17–21).Google Scholar
  41. 45.
    Translated rather ambiguously as “astrologer” by van der Spek (1985) and McEwan (1981).Google Scholar
  42. 46.
    CT 49, 144, Rev. 23; trans. van der Spek (1985: 552).Google Scholar
  43. 47.
    See Rochberg (1999b).Google Scholar
  44. 48.
    See Section 2.3.1 below.Google Scholar
  45. 49.
    Oates (1986: 143).Google Scholar
  46. 50.
    Reade (1986).Google Scholar
  47. 51.
    For details, see Figure 2 of Britton (1993).Google Scholar
  48. 52.
    For similar techniques of dating astronomical papyri, see Jones (1999: v.1 47–55), and horoscopes, see Neugebauer& van Hoesen (1959: 1–2).Google Scholar
  49. 53.
    The Babylonian unit of time was the uš. As there were 360 Uš in one day, it is customary to render Uš as “degree” or ° in translations. See Section 2.4 below for further details.Google Scholar
  50. 54.
    The tablets have been dated at the request of Prof. H. Hunger for their publication in Sachs & Hunger (1999). In all cases, I have worked from his translations of the texts. Further de ta i ls are given in this volume.Google Scholar
  51. 55.
    This criterion is due to Muller (1975), and is a conservative estimate.Google Scholar
  52. 56.
    Steele, Stephenson & Morrison (1997).Google Scholar
  53. 57.
    This map was kindly prepared by Mrs. Pauline Russell and Prof. F. Richard Stephenson of the Department of Physics, University of Durham.Google Scholar
  54. 58.
    See, for example, the instances quoted by Beaulieu & Britton (1994: 77).Google Scholar
  55. 59.
    LABS 347; trans. Parpola (1993: 282).Google Scholar
  56. 60.
    See, for example, Beaulieu & Britton (1994) and Brown & Linsenn (1999).Google Scholar
  57. 63.
    Watches are found, for example, in the omen series Enūma Anu Enlil, see Rochberg-Halton (1988: 44–47); and in the letters and reports sent to the Assyrian king, see Parpola (1993) and Hunger (1992).Google Scholar
  58. 64.
    Neugebauer (1941). 65 Stephenson & Fatoohi (1994).Google Scholar
  59. 66.
    Thureau-Dangin (1937).Google Scholar
  60. 67.
    Oppenheim (1974).Google Scholar
  61. 68.
    Although there exist tables which have been interpreted as referring to measuring time by use of a gnomon, devices such as this could not have been used during the night. 69 See Brown, Fermor, & Walker (1999) for a general discussion of clepsydras in Mesopotamia.Google Scholar
  62. 70.
    M UL. AP I N has been published by Hunger & Pingree (1989), and the relevant parts of Enūma Anu Enlil by Al-Rawi & George (1992).Google Scholar
  63. 71.
    I.NAM.giš has been published by Livingstone (1986).Google Scholar
  64. 72.
    This is mainly due to the surface tension of the water preventing the clock from fully emptying. For the experimental evidence of this statement, see Brown, Fermor, & Walker (1999), and for a theoretical argument, see Høyrup (1998).Google Scholar
  65. 73.
    Historia, II, 109; trans. de Sélincourt (1979: 169).Google Scholar
  66. 74.
    Fotheringham’s reconstruction of the Ivory Prism was published by Langdon (1935), and also discussed by Smith (1969). Fotheringham’s interest in the Ivory Prism was in the possibility of its use in the eclipse records reported by Ptolemy in his Almagest, and so I will defer further discussion of it until Chapter 3.Google Scholar
  67. 75.
    Of course, a seasonal hour is simply a quarter of a watch, but I do not think it was ever extensively used as a unit of time in its own right.Google Scholar
  68. 76.
    Neugebauer & Sachs (1967).Google Scholar
  69. 77.
    Sachs sent this manuscript to F. R. Stephenson, who generously made it available to me.Google Scholar
  70. 78.
    The lunar six data recorded on LBAT 1431 have previously been investigated by Stephenson (1974) who found evidence for systematic clock drifts in the measurement of each of the individual lunar six. He concluded that the Babylonian astronomers may have used a slightly different clock, possibly labeled for the purpose, to measure each of the six intervals.Google Scholar
  71. 79.
    The lunar six measurements may be divided into three groups: those said to have been measured using the term muš, those said to have been not seen using the term NU PAP, and those that have no comment attached. Unless there is some mention of bad weather in the record, I have assumed that the timings in the third category were measured. This may have caused some predicted material to be included in the analysis, but as there is no significant change in the result if this group is ignored, it would appear that, on the whole, these do indeed represent measured lunar six values.Google Scholar
  72. 80.
    The remaining errors are presumably due to the inherent inaccuracies of the clepsydras used for the measurements.Google Scholar
  73. 81.
    See Section 4.3 below.Google Scholar
  74. 82.
    See Hunger & Pingree (1989: 57–67 & 141–144).Google Scholar
  75. 83.
    For these texts, see Schaumberger (1952, 1955) and Horowitz (1994).Google Scholar
  76. 84.
    For an alternative interpretation of the values given in Uš, see Brown (1999a).Google Scholar
  77. 85.
    Lanfranchi & Parpola (1990: no. 249).Google Scholar
  78. 86.
    The first discussion of this terminology was, I believe, by Kugler (1900b). More recently, see Sachs & Hunger (1988: 23).Google Scholar
  79. 87.
    I will discuss this aspect of the Eclipse Texts in Section 2.8 below. A fuller description of the structure of the Eclipse Texts is given in my contribution to Sachs & Hunger (1999).Google Scholar
  80. 88.
    Contrary to the statement by Rochberg-Halton (1989b: 148).Google Scholar
  81. 89.
    These remarks concern not only factors such as clouds that may have affected the observation, but also wind directions which play an important role in the astrological interpretation of the eclipse.Google Scholar
  82. 90.
    BM 36736 has been published by Rochberg-Halton (1984).Google Scholar
  83. 91.
    It is trivial to note that, of course, not all of the observed accounts will include all of the items I have discussed here, and furthermore that many reports are only partially preserved.Google Scholar
  84. 92.
    Unfortunately, due to the vagaries of preservation, there is only a small amount of overlap between years contained on dated Diaries, and those on the dated Almanacs or Normal Star Almanacs. See Hunger (1999).Google Scholar
  85. 93.
    The similarity of the records is even clearer when one considers the texts themselves. LBAT **1059, Obv. 13: 29 1 2 ME ana šamáš AN-KU 1 0 šamáš; LBAT 1151, Obv. 6’ : 29 2 KASGAL-BU ME ana šamáš AN-KU10 fa[más]. Google Scholar
  86. 94.
    Rochberg (1998: 40).Google Scholar
  87. 95.
    A few more reports from this period make reference to ziqpu-stars, but are too damaged to be analysed.Google Scholar
  88. 96.
    The results are reported in Fermor & Steele (1999).Google Scholar
  89. 97.
    In particular he found that with a spout made of steel tubing of bore about 0.68mm and length l 0mm, and a head of water 236mm, the time taken to discharge a fixed weight of water increased by only 6% over the temperature range 30 ° C to 10°C. See Fermor & Steele (1999).Google Scholar
  90. 98.
    Sachs (1974).Google Scholar
  91. 99.
    In his discussion of whether the ziqpu-star timings of eclipses can be used to determine ΔT, Stephenson (1997b: 185) incorrectly dates the observation on 12 July 178 BC to 1 August 188 BC. This is based upon an incorrect reading of the date (SE 124 instead of SE 134) in the copy of LBAT 1439 published by Sachs (1955). The eclipse of 12 July 178 BC was only penumbral and so Stephenson remarks that this is the only known ancient or medieval report of a penumbral eclipse. However, this remark should now be discounted based upon the improved reading of LBAT 1439 by Sachs & Hunger (1999).Google Scholar
  92. 100.
    The identification of stars is a notoriously difficult procedure, and can often not be achieved with full confidence. See, for example, the different identifications proposed by Reiner & Pingree (1981) and Koch (1989). I have accepted the identification of the ziqpu-stars by Sachs and Hunger (1988, 1989, 1996, 1999). 101 See Section 2.4 above. 102 See Section 4.4 below.Google Scholar
  93. 103.
    See, for example, Horowitz (1998: 196–198). I thank C. B. F. Walker for bringing this point to my attention.Google Scholar
  94. 104.
    For a more detailed study of eclipse prediction in Mesopotamia, see Steele (1999a).Google Scholar
  95. 105.
    Koch-Westenholz (1995: 105).Google Scholar
  96. 106.
    See Kugler & Schaumberger (1935: 251), and Rochberg-Halton (1988: 41) whose translation I have largely followed.Google Scholar
  97. 107.
    More generally, this time interval is a very complicated function dependent upon the moon’s longitude and latitude, its velocity, and several other factors. See Neugebauer (1957: 107–110) and Brack-Bernsen & Schmidt (1994).Google Scholar
  98. 108.
    Of course, the definition of the beginning of the Saros period is arbitrary, and the distribution could equally well be 7–8–7–8–8, 8–7–8–8–7, 7–8–8–7–8 or 8–8–7–8–7. The last distribution is that found by Aaboe (1972) from a theoretical analysis of equally spaced eclipse possibilities.Google Scholar
  99. 109.
    In other words, with an eclipse possibility that was a multiple of 223 synodic months before or after the eclipse 6 February 747 Bc.Google Scholar
  100. 110.
    Steele (1999a). 111 Unfortunately, the Diaries for the months on which these eclipses would have been recorded are either missing or damaged and so we do not know exactly which unpredicted eclipses were seen.Google Scholar
  101. 112.
    These texts have been published by Aaboe, Britton, Henderson, Neugebauer, & Sachs (1991).Google Scholar
  102. 113.
    The suggestion by Britton (1993) that the distribution of eclipse possibilities on the Saros Canon represents the scheme that was used to make the predictions of eclipses for the Diaries, and, consequently, that there must have been a reform to the way the Saros was applied from the earlier scheme represented on LBAT *1414 etc. in about 527 BC, can be discounted as the Saros Canon would not predict all of the eclipse possibilities found in the NMAT sources during the period it covers.Google Scholar
  103. 114.
    For a detailed discussion of these various schemes, see Steele (1999a).Google Scholar
  104. 115.
    I follow the terminology established by Neugebauer (1955) in discussing these texts.Google Scholar
  105. 116.
    Aaboe & Henderson (1975).Google Scholar
  106. 117.
    For a detailed discussion of the methods of predicting eclipses using Systems A and B, see Neugebauer (1945, 1955: 41–85, 1975: 474–539).Google Scholar
  107. 118.
    See Section 2.2 above.Google Scholar
  108. 119.
    See Steele (1999a).Google Scholar
  109. 120.
    Sachs & Hunger (1998: 25).Google Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2000

Authors and Affiliations

  1. 1.University of DurhamUK

Personalised recommendations