Pergamon Press Ltd. Oxford·New York·Seoul·Tokyo


The colour of Sirius as recorded in ancient Chinese texts+

JIANG Xiao-yuan
Shanghai Observatory, Chinese Academy of Sciences, Shanghai, 200030


Abstract: At present, Sirius is white, but in the West there has long existed a succession of records purporting that the star is red, which would mean difficulty for the current theory of stellar evolution. In this paper, I have systematically examined the more reliable records regarding the colour of Sirius in extant ancient Chinese texts from B. C. times to the +7th century and found that, during this period, Sirius was unquestionably white, thus relieving the present theory of stellar evolution of any threat from this direction.
Key words: Colour of Sirius-stellar evolution-ancient Chinese texts



  The Heavenly Wolf (ancient Chinese name for Sirius, CM ) is the brightest star in the sky and it is brilliantly white. It is an optical double star, and Star B (the companion) is also the first identified white dwarf. But this famous star has, on account of certain ancient records of its colour, been something of a headache for the current theory of stellar evolution.
  In the western literature, Sirius has often been described as being red. Scholars have found such records in ancient cuneiform tablets of ancient Babylon and in the works of ancient Greek and Roman celebrities including Ptoloemy, Seneca, Cicero and Flaccus. In 1985, schlosser and Bergmann revived the topic, asserting that they had discovered in an early medieval manuscript, a work written by Bishop Gregory of Tours in the +6th century, in which a red star was mentioned which can definitely be identified with Sirius, thus implying that Sirius was still red at the end of the sixth century, and becoming white only after that date [1]. So started a new cycle of debate and concern on the colour of Sirius.
  According to the current theory of stellar evolution and our knowledge about the Sirius binary, Star A is located precisely on the main sequence, and it is impossible for it to change colour in a couple of thousand years. If the binary had indeed a red colour before the +6th century, the only theoretical way out is to shift our attention to the dim Star B, the white dwarf, for before a star evolves into a white dwarf it may pass though a red giant stage. This might offer a hope for explaining the western records, by saying that at that time, the huge red light of Star B blocked out the light of Star A. However, according to the modern theory, the time taken for a red giant to evolve into a white dwarf, even in the extreme case, will be far longer than 1,500 years, so there is yet no satisfactory explanations of these. Western records of Sirius being red.
  Thus, astronomers are faced with the following choice: either to doubt the current theory of evolution, or to deny the proposition that Sirius was red in ancient times.
  The reality of the Western descriptions of the colour of Sirius is by no means unimpeachable. Seneca, Cicero and Flaccus were either philosophers, politicians or poets, and there is no testimony to their astronomical competence; as for Ptolemy, although he was a great astronomer, there is room for doubt on specific details even in his writing. As to the red star described by Bishop Gregory, other scholars [3] have maintained it to be Arcturus ( Boo), which is, of course, a well-known bright red giant.
  In another direction, ancient Chinese astronomical/astrological records are well-known for their abundance, systematic character and fine description of the heavenly phenomena. We should therefore seek further evidence in ancient Chinese texts. To maintain authoritativeness, I shall confine the investigation strictly within dedicated texts, and shall ignore all philosophical or literary writings.



  Astrophysics did not exist in ancient times and ancient people would not have noted the colours of heavenly objects with the modern eyes. Almost invariably, colours of planets and stars were mentioned in ancient texts for their astrological implications. It should be pointed out at the outset that, in the vast majority of cases, such records have no scientific value for the topic under discussion. They usually appear in the same format; here are two examples:

      “... To its east there is a large star called Wolf. If Wolf shows horns (jiao) or changes colour, then there will be much banditry [4].”
      “If the Wolf Star ...... shows rays or horns, or shakes about, or changes colour, then there will be war; if it shows great brilliance, then weapons will be sought after ...... if its colour is yellow and smooth, there will be joy; if the colour is black, there will be sorrow [5].”

    In these quotation, the Wolf Star refers to Sirius (the ancient people did not know it was a double star). Obviously, that the star should change colour, sometimes yellow, sometimes black (other prognosticatory texts also mention red), even undergoing “shaking”, is quite impossible as judged by modern astronomical knowledge. Yet in ancient Chinese astrological writings, these descriptions were applied to many stars (only the prognostications are different in different cases). Even if we force an explanation of “shaking” in terms of illusions caused by atmospheric phenomena in the ancient observers, we must still admit that, for the purpose of finding out the colour of Sirius in ancient times, it is meaningless to resort to records of this kind.
  Fortunately, the ancient Chinese astrological corpus contains another kind of record which, though small in size, is extremely reliable. The ancient people believed not only in the astrological significance of colours of stars, they took the same view of the colours of the planets. The following is an early as well as a most typical statement in this connection:

  “For the five planets, a white halo means funeral or drought; a red halo means disturbance and war in the Middle Kingdom; a blue halo means sorrow and flood; a black halo means disease and many deaths; a yellow halo is auspicious. Red horns mean invasion of our cities; yellow horns, fighting over territories; white horns, sound of walling and weeping; blue horns, war and sorrow; black horns, flood [4] .” 

  That planets should at times change colour and shape is equally impossible, and we can put this aside. But what should be noted is this: since the ancient people believed in the truth of the prognostications, they must have established standards for the colours in fact, their specific method was to take certain famous stars as standards for the various colours. It is necessary to elaborate on this point.
The earliest record in this connection at present available comes from the pen of SIMA Qian. In his discussion of the colour of Venus (Tai-Bai), he gave the five colour standards as follows:

  “For white, compare Lang (Sirius); for red, compare Xin ( Sco); for yellow, compare the Left Shoulder of Shen ( Ori); for blue, compare the Right Shoulder of Shen ( Ori); for black, compare the large star of Kui ( And).”

  The credibility of SIMA Qian's description is confirmed by the following fact: of the five stars, apart from Sirius which, being sub jadice, must be set aside, all the other four were given believable colours, Sco, spectral type M0, is indeed red; Ori, Spectral type B2, is blue; Ori is at present a red supergiant, but it has been established [7] that it is quite consistent with modern theory of stellar evolution for it to have a yellow colour two thousand years ago. The last one, And, spectral type M0, is dark red, but there is justification for the ancient Chinese to call it black. Firstly, in China, the theory of the Five Elements had been current for a long time and had permeated all aspects of thought. Indeed, that stars should be divided into five colours is an important manifestation of the conjoining of this theory with prognostic astrology [8]. But the correspondence between the five elements and five colours had been fixed and the colours must be blue, red, black, white and yellow: there must be a black. Next, since these five colour standards are to be used for comparison in observation, and if it is truly "black", it would be invisible and no comparison would be possible, so the word must not be taken literally.
  As far as the question under discussion is concerned, there is another fortunate circumstance. Since the ancient people took the five colours of the five elements as fixed categories, they must make approximation or accommodation for intermediate states outside and between the five colours, and force such cases into one of the five, therefore when they talked about colours of stars, they cannot help of being inaccurate. Fortunately, this inaccuracy “primary” standards, we need not worry about approximation or accommodation. This circumstance further guarantees reliability of using ancient Chinese texts to solve the question of the colour of Sirius.



  The discussion in the foregoing section shows that only records relating to the five colour standards are reliable. Such records number extremely few in the vast literature of ancient Chinese astrology, yet precisely from these few records we can investigate the colour of Sirius. Table 1 is a survey of all such reliable records that could be found from ancient texts before the +7th century, showing the original text, source, author and date. An analysis and interpretation of these records now follow.

Table 1 Records of the Colour of Sirius from ancient Chinese Texts (100 B.C.-646 A.D.)







as white as Lang1


SIM Qian

100 B.C.


as white as Lang


BAN Gu, BAN Chao, MA Xu

100 A.D.


as white as Lang or Zhinfü2


LIU Biao

200 A.D.


as white as Lang


LI Chun-feng

646 A.D.

1 Lang = Sirius. 2 Zhinfü = Lya.
SJ/TGS = Shiji Tianguanshu. Historical Records, Book of asterisms.
HS]TWZ.= Hanshu Tianwenzhi. History of Hall, Astronomical Chapter.
JZZ = Jinzhou Zhan. Jinzhou Book of Prognostication.
JS/TWZb = Jinshu Tianwenzhi zhong. History of Jin, Astronomical Chapter, Book 2 of 3.

Item 1 It is rather simple. SIMA Qian said of himself: “In the first year of Taichu (104 B.C.), ..., The Astronomer Royal says: ‘my father having spoken, how can I shirk the task ? ’Therefore I began to write” [9]. This, then, is the year when he began to compile Shiji, hence I have roughly assigned the year 100 B.C. as the year for the Book of asterisms. (A precise date of composition is both impossible in practice and meaningless for the present purpose).

Item 2 Hanshu was compiled by BAN Gu (32-92 A.D.), but its astronomical chapter and some other parts were not finished in his life time. This work was done by his sister BAN Chao and MA Xu. This circumstance is recorded in the History of the Later Han:

  “[Her] elder brother Gu wrote Hanshu, but died before he completed its eight tables and the Astronomical Chapter. Emperor He [therefore] decreed that [BAN] Chao should work in the East Library to continue the task .... Later, the emperor decreed [MA] Rong's elder brother Xu to succeed Chao to complete the work [10].”
  “Emperor Xiao-Ming ordered BAN Gu to compile Hanshu, while MA Xu wrote the Astronomical Chapter [12]”
  From the dates of BAN Chao (497-1207 A.D.) and of Emperor He (89-105 A.D.), I have assigned the date 100 A.D..

Item 3 The original text of JZZ is now lost, but it is copiously referenced in KYZJ (Kaiyuan Zhan Jing or The Kaiyuan Book of Prognostication) and YSZ (Yi-si Zhan, or Yi-si Prognostication). It is listed under the name of LIU Biao. LI Chun-feng in his YSZ gave a list of astrological texts which “he learned by heart since a child”. It contained 25 titles, the 18th is LIU Biao's JZZ [l3]. LIY Biao (142-208 A.D.) became the Governor of Jinzhou in 190 A.D., he ruled Jinzhou for a long time, virtually as an independent state. Wheter JZZ came from his pen or the astrologers he gathered about him is not known. Thus, we have to put the record under his name and tie it to the year 200 A.D..

Item 4 It is simple and clear. The History of Jin was completed in 646 A.D., and its “Astronomical Chapter” was written by LI Chun-feng. 

  Although ancient Chinese writings always had a tradition of copying what previous authors had said, so that the four items listed in Table 1 are probably correlated to some degree, we should not therefore deny ancient Chinese astrologers of any originality whatever. That JZZ lists both Sirius and Vega as standards for white star is worth noting. These two are indeed white bright stars of the same type. In the modern MK spectral classification, Sirius is A1V, Vega is A0V, the difference is very small. This further confirms the reliability of the colour of Sirius as recorded in these four records.



  At this point we surely know that, in the ancient Chinese astrological literature, the vast amount of illusory changes of colours of stars and their movements used in prognostication cannot be used for ascertaining the colours of stars at the time, while in the reliable kind of records, Sirius was always white. Not only there has never been any mention of it being red, it was, over the millennia, taken as a standard for white star. This is the case in the early literature examined in this paper, and there has no subsequent change in later times. Hence we can assert that the current theory of stellar evolution is not threatened in any way by the question of the colour of Sirius.
  To conclude, it may be pointed out that, this question not only has puzzled Western astronomers for more than one century, it was introduced into China as early as the last century. In "A History of Astronomy of the Western Countries" (“Xiguo Tianxue Yuanliu”) translated by WANG Tao and A. Wylie in 1890 near the end of the Qing dynasty, we find: "The ancient people invariably said Sirius was red, but to-day it is white, the reason is not known." Unfortunately, scholars have so far failed to trace the original of this book [14]. The significance of this query was not, of course, fully brought into the open then.
  There are some Western scholars who would insist on the correctness of the ancient Western records of Sirius being red and, in an effort to avoid contradiction with modern theory of stellar evolution, therefore proposed hypotheses to the effect that there was a cosmic cloud in the past which covered Sirius and reddened its light [15]. Now that we have the Chinese sources stating unequivocally that Sirius was all the time white, such hypotheses can be excluded once and for all. Further, it may be noted that SIMA Qian predated Ptolemy by some two hundred years and since SIMA Qian took Sirius as standard for white star, it is quite impossible for the star to have changed into red in the interim. It is therefore obvious that Ptolemy's statement that Sirius is red cannot be given credence.


+Supported by National Natural Science Foundation Received 1991 October 22;revised version 1992 May 3

Annotated References

[1] Schlosser, W. and Bergmann, W., Nature, 318 (1985) 45
[2] Since the publication of Ret. [1], there have been, up to 1990, at least six articles on this issue appeared already in the same journal alone. Chinese journals including Exploration of Nature (Vol. 5, No. 2) and Science (1986 No. 6) have also reported on this.
[3] McCluskey, S.C., Nature, 325 (1987) 87, van Gent, R. H., ditto, both believe that what Gregory mentioned was Arcturus, but Schlosser and Bergmann still insisted that it was Sirius two pages on.
[4] SJ-TGS (Historical Records, Book of Asterisms).
[5] LTMY (Secret Garden of the Observatory), originally compiled by YU Ji-cai of the Northern Zhou dynasty, revised by WANG An-Ii et al. of the Northern Song. juan 14.
[6] Gray, C. and Bonnet-Bidaud, J.M., Nature, 347 (1990) 625, is such a case. Based on the statement in SJ that [if the] Wolf [star grows] horns [and] changes colour, [then there will be] much banditry [and] thieving, the authors argued that Sirius was at the time in the course of changing its colour. This is a complete misunderstanding of the real meaning of the text.
[7] BO Shu-ren et al., Kejishi Wenji (Collected Papers on History of Science and Technology) First Issue. (1978) 75-78. Shanghai Kexue Jishu Chubanshe.
[8] JIANG Xiao-yuan, ExpIoration of Nature, 10 (1991) 107.
[9] SJ-TSGZX (Historical Records, Autobiography of the Grand Historian [SIMA Qian]).
[10] HHS-BCZ (History of the Later Han, Biography of BAN Chao, Wife of CAO Shi-shu).
[11] HHS-TWZa (History of the Later Han, Astronomical Chapter, Book 1 of 3).
[12] KYZJ (Kaiyuan Treatise on Prognostication) written by Qutan Xida of Tang. juan 45 reference.
[13] YSZ (Yi-si Book of Prognostication) written by LI Chun-feng of Tang. juan 1.
[14] XI Ze-zong, Journal of Hong Kong University, Chinese Department 1 (1987) No. 2.
[15] Brecher, K., Technology Review 80 (1977) No. 2. The same hypothesis was also maintained in Reft [6].