Appendix IV. Solar Eclipses during the First Fifty Years of the Former Han Dynasty

Since exactly fifty years elapsed from the accession of Kao-tsu in 206 B.C. down to and including the year of Emperor Wen's death, 157 B.C., it may be useful to summarize the results we have obtained in the discussion of the eclipses during that period. Out of the total number of 119 eclipses occurring, only 23 were visible in China. Four of these were so barely visible that we can hardly expect them to have been noticed. Excluding these, only 19 eclipses were visible in Ch'ang-an in this half-century. Six of them were recorded correctly. For three more listings we were able to suggest plausible dates. If these suggestions are accepted, only 9 out of a possible 19 eclipses were recorded--- about half of the total eclipses visible, weather permitting.

In these four reigns, we found 13 records of solar eclipses. Nine referred to actual solar eclipses (accepting the suggested corrections for their dates). One plainly referred to an actual lunar eclipse. In the case of the three remaining we found great difficulties in suggesting any date. For one, we found reason to suspect that it was a later interpolation. For a second, we found that the general opposition on the part of the officials to the Empress Dowager's acts made it possible that an eclipse was fabricated to warn her. The third we found textually doubtful, since Pan Ku had not included it in his "Annals."

The ten plainly visible eclipses that were not recorded were the following: in the reign of Kao-tsu, four: July 6, 206 B.C., Jan. 1, 205, July 26, 197, June 6, 195; in the reign of Emperor Hui, one: Sept. 29, 192; in the reign of the Empress of Kao-tsu, none; in the reign of Emperor Wen, five: Oct. 10, 174, Apr. 4, 173, July 17, 169, May 28, 167, and May 17, 166 B.C. The four others which were barely visible or visible only outside of Ch'ang-an were those of Oct. 8, 201 B.C., May 6, 184, Mar. 26, 164, and Mar. 5, 162 B.C.

Bad weather is a possible cause for the failure to record some of these eclipses. But in Sian (Ch'ang-an), Shensi, during the period 1924-1934, the number of overcast days, as reported by Fr. E. Gherzi, S.J., the Director of the Zikawei Observatory, averaged almost the same for each month in the year, ranging from 7 in July and November to 11 in April and October. The monthly variation of the weather thus cannot account for the omission of these eclipses. Reports could moreover have been secured from places outside the capital.

It is however noticeable that during the years 188 to 175 B.C. all the visible eclipses were recorded, ¹ whereas between 175 and 162 B.C. five eclipses were visible and none were recorded. None of these were total in Ch'ang-an; they might easily have been missed if no one was looking for them. Why were they missed? Was it because of a change in the responsible astronomer, who was negligent in his duties? Or was it because it was considered that the good reign of Emperor Wen left Heaven without any necessity of sending eclipses? Emperor Wen in an edict said that an eclipse was an admonition from Heaven (4: 9a). Neugebauer (op. cit. II, 73) says that an eclipse is not conspicuous unless it reaches a magnitude of at least 0.75, or, at sunrise or sunset, 0.33, so that, unless people were looking for eclipses, they might easily be missed. Yet Halley's comet appeared in 163 B.C. (cf. p. 261, note), and it too was not recorded. Such a spectacle could hardly have been missed. Perhaps comets were not then being regularly recorded; cf. 5: n. 1.3. In later periods of good rule, such as that of Emperor Hsüan, eclipses are similarily lacking in the records.

Several conclusions emerge from a consideration of these eclipses: First, the Chinese accounts are predominatingly reliable, even for the beginning of the Han period. In the later parts of this period, we shall find them much more reliable.

Secondly, there are errors of recording or transmission of the text. One eclipse (the second one in Kao-tsu's reign) is probably a later interpolation; three others were incorrectly dated, possibly before Pan Ku used them. In the case of the lunar eclipse we found positive evidence that an error had been made in transcribing from the original astronomical records.

Thirdly, we found that the court astronomers, in some cases at least, kept records of lunar as well as of solar eclipses, although the historians did not consider them worth recording.

Fourthly, we found that in one case at least (cf. ch. 2: App., ii) astronomical reports were gathered from places outside the capital. The divergences between the list in the "Annals" and that in ch. 27 regarding this eclipse suggest that the list in ch. 27 came from the court astronomers at the capital.

Fifthly, we found that Pan Ku must have had records of eclipses not found in the SC, which he inserted into that material.

Sixthly, eclipses were considered as warnings to the ruler from Heaven, so that during an unpopular reign all visible eclipses were recorded, while during a decade in a "good" reign no eclipses were recorded, not even a conspicuous comet.

Seventhly, our tables of correspondences between European and Chinese dates are quite accurate for Han times, but they may be two days in error.

Eighthly, the records of the position of the sun at the time of an eclipse are not reliable. In the case of the six correctly recorded eclipses, that location is fairly good, but in the only one which was total in Ch'ang-an, that of 181 B.C., in which alone the position of the sun among the stars could have been observed at the time of the eclipse, that position is in error by 14°. In the case of the three eclipses for which we suggested emendations in their date, the location of the sun is grossly in error. In the case of the one which seems a later interpolation, the location of the sun was plainly calculated from the calendar. We suspect that all of these records were so calculated. These locations are moreover not given in the "Annals," but only in the "Treatise on the Five Elements," and in two cases not at all. They are thus of little help in identifying an eclipse.