Did Chinese Records Confirm Jesus’ Crucifixion? The Darkness and Eclipse Claim

Popular posts say Chinese court records of Emperor Guangwu saw the darkness at Jesus’ crucifixion, named a dying “Man from Heaven,” and logged a rainbow three days later. The records are real, but they describe ordinary solar eclipses in AD 30 and AD 31, and a solar halo in AD 31. The Guangwu annals record nothing for AD 32–33. (1) A commentary note reports one more eclipse, most likely that of 12 September AD 33, about five months after Passover. (3) A solar eclipse cannot fall at Passover, and none was visible from Judea in spring AD 33. (2) This page sets out the text, the dates, and what the Gospels themselves claim. What Do Non-Christian Sources Say About Jesus? covers the Western notices.

Key Points

• The Book of the Later Han (Hou Hanshu) records solar eclipses on the last day of the ninth month of Jianwu 6 and the last day of the third month of Jianwu 7, which fall on 14 November AD 30 and 10 May AD 31 (Julian). NASA catalogues solar eclipses on both dates. (1) (2)
• After the AD 31 eclipse, Guangwu left the main hall for five days, blamed his own thin virtue, and later granted an amnesty. The phrase quoted as “the sins of all the people are now on one man” is the emperor’s formula meaning “the fault rests on me alone.” (1)
• “Man from Heaven died” is not in the history’s text. It comes from a commentary note quoting an omen book’s reading for any eclipse on a guihai day. (3)
• The “rainbow three days later” is a solar halo crossed by a white arc on 13 May AD 31, which the treatise reads as an omen of border war. (3)
• Han eclipse records are accurate when checked against NASA’s canon. (4) The Guangwu annals record no eclipse between AD 31 and AD 41, and no earthquake, halo, or rainbow in AD 32–33. (1) A commentary note quotes a solar eclipse report for the seventh month of AD 33, which most likely records the annular eclipse of 12 September AD 33, about five months after Passover. (3) (1) (2)
• Passover falls at full moon, so a solar eclipse is impossible then. Julius Africanus made that objection to Thallus. (5) On 3 April AD 33 there was a partial lunar eclipse. (6) (7)
• The China link began with 17th-century Jesuit readings of this eclipse in the Tongjian Gangmu, a digest of Sima Guang’s Zizhi Tongjian, and the German scholar Theophilus Siegfried Bayer rejected it in 1718. (8) (9)

The Claim as It Circulates

A typical version, on a 2026 blog post, quotes three lines: “The sins of all the people are now on one man,” “In the day of Gui Hai, Man from Heaven died,” and, “three days after the eclipse,” a “halo–a rainbow” around the sun. (10) It sets them beside Mark 15:33Mark 15:33 · King James VersionAnd when the sixth hour was come, there was darkness over the whole land until the ninth hour.Mark 15:33 · King James VersionOpen source (11) and treats the halo as a sign of the resurrection. The same post dates the eclipse to around AD 32. (10)

The idea is old. Seventeenth-century Jesuit missionaries found this eclipse through the Tongjian Gangmu, a digest of Sima Guang’s Zizhi Tongjian (1084), not through the Book of the Later Han. (8) The Zizhi Tongjian entry is short: 「癸亥晦，日有食之。詔百僚各上封事，其上書者不得言聖」 (“On the last day of the month, guihai, the sun was eclipsed. The officials were ordered to submit sealed memorials, and writers were not to call the emperor a sage”). (15) Philippe Couplet’s chronology in Confucius Sinarum Philosophus (Paris, 1687) recorded it with a note that astronomers would have to decide whether it was the eclipse at Christ’s death. (8) In 1718 Bayer published De Eclipsi Sinica, which examined the claim that this was the eclipse at the crucifixion. (9) He argued that the Chinese entry is one routine eclipse among many, that no eclipse lasts three hours, that Passover falls at full moon, and that the word sheng (聖, “sage”) in the edict refers to the emperor, not to Christ. (8)

What the Han Records Say

The annals of Emperor Guangwu (Hou Hanshu, juan 1B) and the Five Phases treatise (Hou Hanshu, treatise 18) carry every entry the claim uses. The translations here are this page’s own. Julian dates come from the sexagenary day names, which match NASA’s eclipse dates. The Wikisource edition of the treatise prints some passages in simplified characters, and they are quoted here as printed. (1) (2) (3)

The day after guihai is the first of the fourth month, so bingyin is three days after the eclipse. That is the real basis of the “three days later” line. The record describes a halo display around the sun, not a rainbow after rain, and it is dated AD 31.

In the edict, 予 is the first-person pronoun, and yu yi ren (予一人, “I, the one man”) is the emperor speaking of himself. The sentence takes the people’s faults on the ruler, as the earlier edict blames his “thin virtue.” Reading 予 as “give” and 一人 as another man produces the claimed “sins … on one man,” which the grammar does not support. (1)

The annals for Jianwu 8 and 9 (AD 32–33) cover the campaign against Wei Xiao, floods, and appointments. They record no eclipse, earthquake, halo, or rainbow. The next solar eclipse in the annals is in Jianwu 17 (AD 41). (1)

The treatise commentary does quote the Gujin Zhu for three later eclipses: 「九年七月丁酉，十一年六月癸丑，十二月辛亥，並日有蝕之」 (“Ninth year, seventh month, dingyou; eleventh year, sixth month, guichou; twelfth month, xinhai: on each the sun was eclipsed”). (3) The annals record a leap month in Jianwu 8, so the seventh month of Jianwu 9 ran from about 14 August to 12 September AD 33 (Julian). (1) That month has no dingyou day. Its last day, 12 September AD 33, is the annular solar eclipse NASA lists as number 04857. (2) The entry most likely records that eclipse with a corrupted day name; copyists’ errors do occur in Han eclipse reports. (4) Either way it is about five months after Passover and does not support the claim.

A commentary note in the treatise does cite a lunar halo of five rings, “colored like a rainbow,” on a night in the third month of Jianwu 8 (AD 32). It is a nighttime halo, not daytime darkness, and no version of the claim uses it. (3)

Why a Solar Eclipse Cannot Be the Passover Darkness

A solar eclipse happens only at new moon, when the moon passes between the earth and the sun. Passover begins on 14 Nisan, at full moon, when the moon is opposite the sun. Africanus made this point against Thallus: an eclipse cannot happen “when the moon is almost diametrically opposite the sun.” (5) A total solar eclipse also lasts minutes at any one place, not three hours. (2)

NASA’s Five Millennium Canon lists solar eclipses on 19 March and 12 September AD 33, neither at Passover and both with greatest eclipse far from Judea (40°S and 53°N). (2) The Han eclipses of AD 30 and AD 31 also fell on the last day of a lunar month, which is the new moon. The Chinese entries are themselves evidence that these were not full-moon events.

The Gospels do not call the darkness an eclipse. Mark 15:33Mark 15:33 · King James VersionAnd when the sixth hour was come, there was darkness over the whole land until the ninth hour.Mark 15:33 · King James VersionOpen source says “there was darkness over the whole land until the ninth hour.” (11) Luke 23:44–45Luke 23:44–45 · King James Version^23:44 And it was about the sixth hour, and there was a darkness over all the earth until the ninth hour. ^23:45 And the sun was darkened, and the veil of the temple was rent in the midst.Luke 23:44–45 · King James VersionOpen source adds that “the sun was darkened.” (12) Matthew 27:51Matthew 27:51 · King James VersionAnd, behold, the veil of the temple was rent in twain from the top to the bottom; and the earth did quake, and the rocks rent;Matthew 27:51 · King James VersionOpen source records that “the earth did quake, and the rocks rent.” (13) Christians have read the darkness as an act of God. Africanus calls it “a darkness induced by God.” (5)

The Lunar Eclipse of 3 April AD 33

NASA lists a partial lunar eclipse on 3 April AD 33 (Julian), umbral magnitude 0.576, with greatest eclipse at 17:37 Terrestrial Time and an umbral phase of about 170 minutes. (6) With NASA’s ΔT of 10,202 seconds, greatest eclipse was about 14:48 Universal Time, late afternoon in Jerusalem.

Colin Humphreys and Graeme Waddington used this eclipse in Nature (1983). Their abstract says they dated “a lunar eclipse that biblical and other references suggest followed the Crucifixion,” and that the evidence points to Friday 3 April AD 33. (7) A lunar eclipse cannot darken the sun, so it does not explain a darkness from the sixth hour to the ninth. Friday 7 April AD 30 remains the other date many historians defend, and neither date depends on a Chinese record.

What Thallus and Phlegon Claimed

Neither writer mentions China. Thallus’s history is lost. Julius Africanus says that Thallus, in the third book of his history, called the darkness an eclipse of the sun, “as appears to me without reason.” (5) Africanus also reports that Phlegon “records that, in the time of Tiberius Caesar, at full moon, there was a full eclipse of the sun from the sixth hour to the ninth.” (5)

Origen says Phlegon wrote about an eclipse in the reign of Tiberius “and the great earthquakes which then took place” in the thirteenth or fourteenth book of his Chronicles. (15) Phlegon’s own words survive only in these later quotations. These are early reports that someone outside the church discussed a darkness and earthquakes. They do not confirm a Chinese observation.

Conclusion

The Chinese records are genuine and precise. Guangwu did react to the solar eclipse of 10 May AD 31 with self-reproach and an amnesty, and a solar halo was recorded three days later. Those events took place two years before 3 April AD 33, the one Han observation that names a dying “heavenly man” is an omen formula for any eclipse on that cyclical day, the Guangwu annals have nothing for AD 32–33, and the one commentary report for AD 33 most likely records the eclipse of 12 September, about five months after Passover. Han China gives no independent witness to the crucifixion darkness or earthquake. The Gospels present that darkness as a sign from God, not as an eclipse anyone could predict, and the case for the crucifixion rests on the sources in Did Jesus Really Exist? and The Resurrection of Jesus.

Frequently Asked Questions

Did Chinese Astronomers Record the Darkness When Jesus Died?

No. The nearest records are solar eclipses on 14 November AD 30 and 10 May AD 31. The Guangwu annals record no eclipse or other sign in AD 32 or AD 33. (1) A commentary note reports an eclipse in the seventh month of AD 33. It most likely records the annular eclipse of 12 September AD 33, about five months after Passover. (3) (2)

Does the Book of the Later Han Say “Man From Heaven Died”?

Not in its own text. The phrase translates an omen-book line, “an eclipse on a guihai day: a heavenly man dies,” quoted in Liu Zhao’s commentary on the treatise. It is a stock reading for the day name, not a report of a death. (3)

Was There a Rainbow Three Days After the Crucifixion?

The record is a solar halo crossed by a white arc on 13 May AD 31, three days after that year’s eclipse. It has no link to April AD 33. (3)

Could a Solar Eclipse Have Caused the Three Hours of Darkness?

No. Passover is at full moon, and solar eclipses happen only at new moon. A total eclipse also lasts minutes, not hours. (5) (2)

Keep Reading

• What Do Non-Christian Sources Say About Jesus?
• Did Jesus Really Exist?
• The Resurrection of Jesus
• Luke 23