← Back to 宋書

卷十二 志第二 歷上

Volume 12 Treatises 2: Calendar 1

Chapter 12 of 宋書 · Book of Song
← Previous Chapter
Chapter 12
Next Chapter →
1
耀 使 使
Life is what Heaven and Earth prize most; among the myriad creatures, it is humanity that commands the highest regard. By applying intellect and pushing the mind to its limits, nothing concealed lies beyond notice; thus every deed and utterance answers to the patterns of Heaven and Earth. The sage rulers of antiquity took the celestial pole as their model and devised the armillary sphere. Yin and yang—the two vital breaths—shape and nourish all living kinds; the luminous bodies in which their essence is embodied are the sun and the moon. The character of all living things displays itself in the five endowments, and the animating power of those five endowments resides in the five planets. The calendar exists to mirror the course of Heaven, set the seven luminaries in sequence, register the affairs of the myriad realms, and grant humankind its proper seasons. The Yellow Emperor charged Da Rao with devising the six jia cycle; Rong Cheng established the calendar and its celestial models; Xi He took charge of observing the sun, and Chang Yi of observing the moon. Under Shaohao the phoenix appeared as a favorable omen; he named his ministries after birds, and the Ministry of the Phoenix Bird oversaw the calendar. In Zhuanxu's time Chong, rectifier of the south, governed affairs of Heaven, while Li, rectifier of the north, governed affairs of Earth. Yao in turn raised up the line of Chong and Li, restored them to their former offices, and separately appointed Xi and He with the charge to honor the great Heaven. Hence the Yu section of the Documents states: "The full cycle is three hundred sixty-six days and a fractional sixth; intercalary months establish the four seasons and complete the year." He then passed the charge to Shun, saying, "Heaven's calendrical mandate is vested in you alone." Shun in turn entrusted the same charge to Yu. Through the Yin and Zhou dynasties alike, each new house founded its rule and revised its institutions, and court ritual colors followed suit. By aligning with the seasonal breaths they answered Heaven's Way, and the myriad creatures reaped the benefit of that harmony. Once the Three Kings were gone, the historiographical service lapsed and its posts stood empty; for this reason Confucius rectified the Spring and Autumn to lay bare the errors of those charged with the calendar. When Qin united the empire, it claimed the Water phase as its own, took the tenth month as the year's beginning, and elevated black in ritual dress.
2
調調 使 調[1]
At the founding of Han the dynasty retained Qin's calendar and new year's day; Zhang Cang, Marquis of Beiping, was the first to address matters of pitch-pipes and the calendar, comparing the Zhuanxu system with the six rival calendars—and the discrepancies proved almost equally slight. This system remained in use until the seventh year of Yuanfeng, when Gongsun Qing, Hu Sui, Sima Qian as Grand Astrologer, and others declared that the calendrical cycle had broken down and that the new year's month and ritual colors should be revised to show plainly that the mandate came from Heaven. The emperor accordingly ordered Sui and his colleagues to draft the Han calendar. Deng Ping, Director of Changle, Sima Ke, and more than twenty calendar specialists drawn from the general population were chosen for the task. The adept Tang Du apportioned the celestial regions, while Luoxia Hong performed the calculations that set the calendar in motion. By that method, eighty-one fractional parts accumulate to constitute one day's fractional share. Luoxia Hong and Deng Ping worked from the same principles. They then observed stellar positions and the courses of the sun and moon; further reckoning confirmed Hong and Ping's formula—a lunar month comprised twenty-nine days and forty-three eighty-firsts of a day. An edict directed Sima Qian to adopt the eighty-first-part pitch-pipe calendar that Ping had devised, and Ping was appointed Assistant Grand Astrologer. In the third year of Yuanfeng, Grand Astrologer Zhang Shouwang memorialized the throne, arguing that the inaugural year had employed the Yellow Emperor's tuning calendar: "Yin and yang are now out of harmony because the calendar was changed." The court ordered the Chief Calendar Officer Xianyu Wangren, Assistant Director of Agriculture Ma Guang, and more than twenty colleagues jointly to observe new moons, full moons, quarter moons, and the twenty-four seasonal nodes. A further edict assigned one clerk each from the chancellor, imperial secretary, grand marshal, and right general to observe at the Clear Terrace in Shanglin and grade every system for tightness or slack—eleven schools in all. Observations began in the third year and ran through the fifth. Shouwang's calendar ranked lowest for being loose and far off the mark. Moreover, Han's inaugural year had not in fact used the Yellow Emperor's tuning calendar; Shouwang was charged with defying Heaven and Earth—a grave offense of supreme irreverence. An edict forbade proceeding with the impeachment. Observations were resumed through the sixth year, and the Taichu calendar placed first. Shouwang's system turned out to be the Yin calendar preserved in the Grand Astrologer's archives. When Shouwang was impeached a second time and refused to yield, he was at last remanded to the magistrates. Under Emperor Cheng, Liu Xiang surveyed the six calendars, weighed their merits and faults, and wrote the Treatise on the Five Cycles. Xiang's son Xin devised the Triple Concordance calendar to interpret the Spring and Autumn; for all its ingenious phrasing and parallel structures, it did not accord with fact. Ban Gu pronounced it subtle and essential, which is why the Han Treatise on the Calendar preserves it. Set against the calendars of He Tiancheng and five other schools—though their six origins differed and their fractional schemes diverged—the accumulated error down to the present runs to three days, or two days and several hours; judged by antiquity, nearly all were the work of scholars from the Six States and Qin. Their methods employ oversized dipper fractions: they cannot be tested against the Spring and Autumn in earlier times nor confirmed against Han and Wei in later ones; even when they falsely attribute themselves to ancient emperors, they are fit only to delude the men of their own day.
3
宿宿 [2] [3]
In the eighth year of Jianwu under Emperor Guangwu, Grand Minister Zhu Fu submitted a memorial stating that the calendrical cycle was wrong and should be corrected. The error was still small, and the court had no leisure to undertake a full correction. Under Emperor Ming in the Yongping period, Yang Cen, Zhang Sheng, Jing Fang, and other awaiting-edict scholars oversaw the calendar but merely adjusted the hours at which quarter moons were marked; they could not carry out a full verification of the calendrical origin. By the second year of Yuanhe the Taichu system's drift from Heaven had grown pronounced; stellar lodge positions showed a steadily widening gap; every observer knew the sun's lodge was five degrees in error, winter solstice stood in Dipper 21, and new moons, full moons, and quarters appeared a day too early. Emperor Zhang summoned the calendar experts Bian Xin, Li Fan, and others to conduct a comprehensive review. He then issued an edict citing the Spring and Autumn and the Baojian tu: "Every three hundred years the Dipper calendar must revise its standard. The court astronomers still employ the Taichu method of Deng Ping, which retains a surplus fraction of one; over the three-hundred-year span its computed motion drifts ever further astray; the sighting tube no longer points true, and the celestial patterns fail to match. On winter solstice the sun stood in Dipper 21—one day before Establishment of Spring, which in the Quarter-Day calendar is itself the day of Establishment of Spring. Yet to settle lawsuits and pronounce capital sentences when the seasonal breath had already reversed course— to take the full moon as the moment of evenhanded justice was equally wide of the mark. We now adopt the Quarter-Day system to follow Yao, accord with Confucius, uphold Heaven's pattern, and with one accord reverently grant the seasons—perhaps thereby attaining universal radiance." The Quarter-Day calendar was thereupon put into effect. Every calendar since the Yellow Emperor that had placed winter solstice at the start of the Ox constellation was set aside.
4
In the fourteenth year of Yongyuan under Emperor He, Grand Astrologer Huo Rong, then an awaiting-edict, wrote: "The official water clock gains or loses one notch every nine days on average, failing to match Heaven; at times the discrepancy reaches two and a half notches—it is less exact than the Xia calendar." In the eleventh month of that year, on the day jia-yin, an edict declared: "The clepsydra exists to regulate the hours and fix dusk and dawn. The varying length of dusk and dawn stems from how far the sun stands from the pole; the sun's path is circular and cannot be apportioned by simple proportional calculation. The official practice of adding or subtracting one notch every nine days misses the reality; deriving the notches from gnomon shadows is tighter and can be verified. Let the gnomon-based clepsydra of forty-eight arrow-tubes be issued." The sun's lodge on each of the twenty-four qi, together with the yellow path's polar distance, gnomon shadow, clepsydra gradations, hours of dusk and dawn, and midnight culminating stars, are all tabulated in the Continued Han Treatise on Pitch-pipes and the Calendar.
5
[4] 西
In the third year of Yanguang under Emperor An, Palace Attendant Xuan recited a memorial arguing for the jia-yin origin, while Liang Feng of Henan urged a return to the Taichu system. Secretarial Gentlemen Zhang Heng and Zhou Xing were both expert calendarists; they repeatedly cross-examined Xuan and Feng by calculation, and sometimes the latter could not answer, or their figures were shown to be wrong. Heng and his colleagues reviewed the observatory regulations, compared past with present, and judged the Nine Paths method the most precise. An edict referred the matter to the chief ministers for detailed deliberation. Grand Marshal Kai and others reported jointly: "The Taichu runs one degree ahead of Heaven, and the moon is seen in the west on the last day of the month. Under Yuanhe the court switched to the Quarter-Day system; although the Quarter-Day is tighter than the Taichu, it too is in error. Neither may be adopted. The jia-yin origin accords with Heaven and agrees with the prognostic texts—it should be put into practice." The disputants could not agree. Secretarial Director Zhong memorialized: "Heaven's calendrical numbers must not be left to doubt and groundless speculation, substituting error for truth." Xuan and his party accordingly let the proposal drop.
6
退 耀宿
In the fourth year of Xiping under Emperor Ling, Palace Gentleman Feng Guang and Chen Huang, the Pei commandery reporting officer, declared: "The calendrical origin is wrong, and that is why robbers and rebels inflict harm. They argued that the calendar should adopt jia-yin as its origin rather than geng-shen, and requested the canonical and weft passages underlying the geng-shen origin." An edict referred the question to the Three Excellencies and to Confucian scholars versed in the Way and its arts for thorough debate. The officials gathered at the Minister of Education's headquarters to debate jointly. Consulting Gentleman Cai Yong declared: "Calendrical reckoning is exquisitely subtle; no method remains permanently correct. At Han's founding the dynasty took over Qin's institutions; it employed the Zhuanxu calendar with the yi-mao origin. One hundred and two years later Emperor Wu first adopted the Taichu calendar, with the ding-chou origin. That system remained in use for one hundred eighty-nine years until Emperor Zhang switched to the Quarter-Day calendar with the geng-shen origin. Guang and his allies now declare geng-shen wrong and jia-yin correct. By calendrical reckoning, the Yellow Emperor, Zhuanxu, Xia, Yin, Zhou, and Lu each possessed its own origin. The origin that Guang and Huang invoke is that of the Yin calendar. From the first adoption of the Taichu ding-chou origin, six rival schools quarreled without resolution, contesting right and wrong. Zhang Shouwang had wielded the jia-yin origin against the Han calendar; joint observations at the Clear Terrace ranked his system last. The Taichu system's empirical tests showed no gaps or failures. Thus, although it was not the origin favored by prognostic literature, it had proved effective in its own day. Moreover, since the Quarter-Day calendar came into use, examination of celestial motion has shown it tighter than the Taichu—the newer origin has again proved effective in our own time. That is why in the Yanguang period Xuan's memorial likewise rejected the Quarter-Day and urged the jia-yin origin—yet after joint deliberation by the chief ministers, the change was never carried out. Besides, the courses of the sun, moon, and planets—now slower, now swifter, advancing and retreating—need not be uniform. Hence there are methods suited to antiquity and methods suited to the present. Methods of today cannot be extended upward to antiquity, just as ancient methods cannot be extended downward to the present. Furthermore, Guang and Huang treat the Kaoling yao as their authority, yet the lodge-degree figures for the sun's daily position contain irreconcilable discrepancies. The Quarter-Day calendar adopted in Yuanhe 2 has been in use ninety-two years, yet Guang and Huang blame every disharmony of yin and yang and every traitor and thief on the calendrical origin. The Yuanhe edict is complete in wording and firm in intent; it is not a matter that debating ministers may overturn." The Three Excellencies sided with Yong; Guang and Huang were judged irreverent and sentenced to convict labor under the ghost-labor statute. An edict forbade imposing punishment.
7
He Tiancheng remarked: In the art of calendrical reckoning, when the mind fails to comprehend the method, even the keenest experts cannot keep the system from breaking down. That is why, across many years and generations, no lasting standard has been settled. Measured against Heaven, the Quarter-Day calendar accumulates one full day's error every three hundred years. Age after age failed to notice the drift, insisting that every new calendar must begin by fixing an origin, invoking prognostic texts to tie calendrical choice to the rise and fall of states—a delusion already carried to extremes. Liu Xin's Triple Concordance system is especially loose; set beside the Quarter-Day, it accumulates yet another day's error every six thousand years or so. Yang Xiong was beguiled by the system and wove it into his Taixuan; Ban Gu pronounced it the most exact and entered it in the Han Treatises. Sima Biao then wrote, "The Triple Concordance calendar came into use in the first year of Taichu and remained in force for more than a century. Yet he forgot that Liu Xin was born only after the Taichu reform—were these worthies, one wonders, debating the calendar in near-total ignorance?
8
使
During Guanghe, Liu Hong of Gucheng Gate first saw that the Quarter-Day calendar no longer hugged the sky; he reset the era divisor to 589 and the dipper fraction to 145, created the Qianxiang system, and added a slow-fast lunar ephemeris. Set beside the Taichu and Quarter-Day systems, it marked a clear advance in precision. Early in Wei Wendi's Huangchu reign, Deputy Astrologer Han Yi argued that Qianxiang had pared the dipper fraction too sharply and would eventually run ahead of the sky; he therefore drafted the Huangchu calendar with an era divisor of 4883 and a dipper fraction of 1250. Minister Chen Qun then submitted a memorial: "Calendar theory is notoriously opaque, and fine scholars of past dynasties rarely agreed. The Huangchu reform began from the premise that the old Quarter Remainder had grown hopelessly slack; now that Wei held the mandate, the seasons needed a new calendar. Han Yi opened the discussion, yet his work still lacked rigorous vetting, so the Qianxiang system was brought in for a side-by-side test. For three years the rival camps traded contradictory verdicts, quibbling over minutiae while missing the main issue, disputing measurements without end—and never reached a settlement. The three high ministers replied that every proposal drew on solid classical principle and aimed at the same end by different routes; the thing to do was to run each system on the armillary instrument for a full year—then success and failure would show plainly, which fit the practical need." The throne approved the memorial. Under Emperor Ming, Secretariat official Yang Wei produced the Jingchu calendar, which remained in use through Jin and Song. The great calendar makers of old each advanced the art in a different way: Deng Ping renewed inherited practice; Liu Hong first trimmed the quarter-day excess and modeled the moon's variable speed; Yang Wei split the difference between rival systems, then built correction tables from accumulated conjunction data to predict new moons and eclipses. These three were the foremost calendrical experts of Han and Wei. Yet Liu Hong's lunar tables fail when tested against the Spring and Autumn, and Yang Wei's planetary positions drift badly in later times—Hong's approach was still insufficiently rigorous, and Wei was hamstrung by forcing every calculation back to the shared renchen upper origin.
9
In Jingchu 1 under Wei Emperor Ming the court reset the calendar: the chou month became month one, and what had been the third month was relabeled early summer, the fourth month. Seasonal names no longer matched the Xia calendar, yet for suburban rites, qi-welcoming ceremonies, temple offerings, seasonal sacrifices, royal tours and hunts, the four turning points of the year, and the publication of seasonal edicts, the yin month still counted as the true pivot of the year. When the emperor died in the first month of the third year, the court reverted to the Xia-style first month.
10
西 [5] 使 使 使
Yang Wei wrote: "I have combed the sources and worked through the calendar's numbers: seasons organize agriculture, months organize administration—and that arrangement goes back to remote antiquity. Already under Shaohao the swallow asterism marked the equinoxes; under Zhuanxu and Di Ku, Chong and Li were charged with celestial regulation; under Yao and Shun, Xi and He took charge of the solar calendar. The three early dynasties kept the office, so every reign maintained specialists charged with the calendar. Those officers set the calendar and issued it to the regional lords, who in turn published it inside their own territories. Later in the Xia era, Xi and He neglected their duty and deranged the calendar, which is why the Book of Documents preserves the punitive campaign against them. From this it is clear that every dynasty has treated careful timing for agriculture and attention to civic order as a standing principle. By late Zhou the realm fractured: monthly temple announcements lapsed, and observatory ceremonies died out. Leap months drifted out of place, the civil year's first month wandered, and while Antares still hung in the western sky, courtiers marveled at insects that failed to burrow—blind to the calendar's chaos. The king no longer coordinated the seasons, official scribes stopped reliably dating events, regional rulers ignored their responsibilities, astronomers lost track of true conjunctions, and governance neglected both people and the agricultural timetable. Confucius used the Spring and Autumn to restore order through moral judgment: he mocked missed intercalations, yet praised rulers who climbed the terrace to proclaim the civil calendar as behaving ritually. Down through Qin and Han the year again began in early winter, leap months were tacked after the ninth month, solar terms and lunar months slipped badly out of alignment, ephemeris corrections ran behind the sky, eclipses missed conjunction, and the mistake persisted reign after reign [5] without overhaul. Only in the seventh year of Yuanfeng did Emperor Wu recognize the fault. He reformed the civil calendar, commissioned the Taichu system, adjusted the leap fraction against true conjunctions, checked stellar longitudes, adopted jianyin as the first month of the year, and anchored the epoch in the yellow-bell month. The Taichu system's fractional remainder for the Dipper was too big, so in later reigns it grew increasingly imprecise. From Yuanhe 2 the Quarter-Day calendar returned and was put back into use. Eclipse records down to our own day show conjunctions drifting to month-end—proof the fractional excess was too large, yielding an apparently tight fit that soon loosened into uselessness. Hence I reworked the official surplus-day reckoning, checked it against eclipses and prior canons, and drew up a denser calendar that tracks the sky evenly—neither ahead nor behind—between antiquity and today. This matches Yao's age, when aligning the days and seasons allowed every office to function and every undertaking to thrive. He wants current state ritual and every institution to echo ancient models; so he resets the civil year, revises the calendrical constants, begins the year in the great-lü month, and sets the computational epoch in the jianzi month. I note that each great reform named its system—the Zhuanxu method in the age of the emperors, the Yellow Emperor's calendar from Xuanyuan onward. Han Emperor Wu likewise reformed the new year, revised the constants, and took the era name Taichu—giving the system its title. Since the reign era is now Jingchu, the new system ought to be called the Jingchu calendar. The Jingchu scheme I propose keeps the parameters lean yet numerically tight, runs efficiently in practice, and is straightforward to learn. No amount of legendary arithmetical skill—merchant Yan's mental sums, Lishou's rods, Chong and Li at the sundial, Xi and He at the solstice shadow—could match the precision I have reached. That is why earlier dynasties' calendars always erred on the loose side, and why reform has followed reform ever since the Yellow Emperor.
11
From the Gengchen origin through the Jingchu era's first year (a Dingsi year in the cycle), the accumulated count is 4046, inclusive on the upper stem count. The epoch uses the true winter month (jianzi) aligned with the yellow-bell pitch month as calendar zero, so at the origin year's start a Jiazi day begins at midnight exactly at winter solstice.
12
The origin divisor (yuanfa): 11,058.
13
The era divisor (jifa): 1,843.
14
Months per full era cycle (jiyue): 22,795.
15
Rule years (zhang sui): nineteen.
16
Rule months (zhang yue): 235.
17
Rule intercalations (zhang run): seven.
18
The communication number (tong shu): 134,630.
19
The day divisor (ri fa): 4,559.
20
The remainder number (yu shu): 9,670.
21
The circuit-of-heaven constant (zhou tian): 673,150.
22
[6]
Mid-year qi [6]: twelve.
23
The qi divisor (qi fa): twelve.
24
The extinction fraction (mo fen): 67,315.
25
The extinction divisor (mo fa): 967.
26
The lunar circuit constant (yue zhou): 24,638.
27
The communication divisor (tong fa): forty-seven.
28
[7]
The conjunction communication constant (hui tong): 790,110. Editorial note 7.
29
The new- and full-moon combined number (shuo wang he shu): 67,315.
30
The node-crossing limit number (ru jiao xian shu): 722,795.
31
The communication circuit constant (tong zhou): 125,621.
32
The circuit-day remainder (zhou ri ri yu): 2,528.
33
The circuit void (zhou xu): 2,031.
34
The dipper fraction (dou fen): 455.
35
Era series I: Jiazi.
36
At the initial conjunction of the era the moon is north of the ecliptic.
37
The eclipse-node difference rate for this era: 412,919.
38
The lunar anomaly difference rate: 103,947.
39
Era series II: Jiaxu.
40
At this era's starting conjunction the moon is inside the solar track.
41
The conjunction-node difference rate: 516,529.
42
The anomaly difference rate: 73,767.
43
Era series III: Jiashen.
44
At era head the moon lies inside the ecliptic at conjunction.
45
The conjunction-node difference rate: 621,139.
46
The anomaly difference rate: 43,587.
47
Era series IV: Jiawu.
48
At the era-head conjunction the moon remains inside the solar path.
49
The conjunction-node difference rate: 723,749.
50
The anomaly difference rate: 13,407.
51
Era series V: Jiachen.
52
[8]
At the era-head conjunction the moon is on the inner side of the sun's path. Editorial note 8.
53
The conjunction-node difference rate: 37,249.
54
The anomaly difference rate: 108,848.
55
Era series VI: Jiayin.
56
At the era-head conjunction the moon lies inside the solar track.
57
The conjunction-node difference rate: 140,859.
58
The anomaly difference rate: 78,668.
59
滿 滿 滿滿
The step from one era's conjunction-node rate to the next: 103,610. To derive it: take the months in one era, multiply by the communication number, discard full multiples of the conjunction communication constant; what remains defines the era increment. Add that increment to the previous era's rate to get the next era's rate. If the sum stays below the conjunction communication constant, the new moon at the era's first civil year lies inside the solar track. If it crosses and you subtract the product, the moon at conjunction lies outside the solar track. Stepping from outside, a full increment places the moon inside; stepping from inside, a full increment places it outside.
60
The slow-fast era difference: 30,180. To derive it: multiply one era's months by the communication number, reduce modulo the communication circuit constant, subtract the remainder from that constant; what is left is the era increment. Subtract that decrement from the prior era's anomaly rate to obtain the next. If you cannot subtract, add the communication circuit constant first.
61
For the next grand origin, subtract the previous Jiayin-era rate from the Jiazi-era rate as prescribed to get the new Jiazi increment. To advance to the following era, repeat the same rule.
62
To count lunations from the Renchen epoch up to but not including the target year, divide by 1,843: the quotient plus one gives the era index, the remainder the year within that era. Multiply that year-count by 235 and divide by 19 for total months; the remainder is the leap fraction. A leap year occurs when the intercalary remainder reaches twelve or higher. The leap month is the lunar month that contains no major solar term.
63
滿
To compute conjunction instants, multiply accumulated months by 134,630; divide by 4,559 for whole days; the residue is the fractional day. Reduce the day count modulo sixty for the sexagenary stem-branch index. Index that remainder against the sexagenary cycle to name the day of the eleventh month's conjunction in the target civil year. For the following month add 29 to the day count and 2,419 to the fraction, carrying overflows by 4,559 into the day count, then read off the next conjunction. If the fractional part is 2,140 or greater, the month is a long thirty-day month.
64
滿滿滿 [9]
For quarters: add 7 days, 1,744 parts, and one small part to the new-moon line, propagating carries through the small-part, fractional-day, and sexagenary levels; the result names the first-quarter day. Repeat the same addition chain to reach full moon, third quarter, and the next conjunction. When a lunar eclipse falls on the full moon, adjust the fractional day by the gap limit of the nearest major term; if the value lies at or under the limit number, per note [9], round upward to fix the day. When full moon lies within four days before or after the major term, use the standard limit figure. When full moon sits five or more days away from the major term, switch to the wider gap-limit criterion instead.
65
滿 滿 滿滿[10]
To place the twenty-four solar terms: take years elapsed within the current era, not counting the target year, multiply by 9,670, divide by 1,843 for the day count, and keep the fractional remainder. Reduce the day line modulo sixty and read it against the sexagenary cycle to name the winter solstice of the eleventh month. For each successive term add 15 days, 402 parts, and 11 minute-fractions, carrying overflows as prescribed; per note [10], index the result as before to name the next solar term.
66
滿滿 退
To find leap months: subtract the leap fraction from 19, multiply the residue by 12, divide by 7, and add an extra leap month whenever the remainder reaches half the divisor. Count forward from the eleventh month; the slot so reached is the intercalary month. Leap placement can shift forward or back, but the governing rule remains that no major solar term may fall within the leap month.
67
Major Snow, eleventh month, minor term. 〈Limit number: 1,242. Gap limit: 1,248.〉
68
Winter solstice, eleventh month, major term. 〈Limit number: 1,254. Gap limit: 1,245.〉
69
Minor Cold, twelfth month, minor term. 〈Limit number: 1,235. Gap limit: 1,224.〉
70
Major Cold, twelfth month, major term. 〈Limit number: 1,213. Gap limit: 1,192.〉
71
[11]
Beginning of spring, first month, minor term. 〈Limit number: 1,172. Gap limit: 1,147.〉 Editorial note 11.
72
[12]
Rain water, first month, major term. 〈Limit number: 1,122. See note [12]. Gap limit: 1,093.〉
73
[13]
Waking of insects, second month, minor term. 〈Limit number: 1,065. Gap limit: 1,036.〉 Editorial note 13.
74
Spring equinox, second month, major term. 〈Limit number: 1,008. Gap limit: 979.〉
75
Clear and Bright, third month, minor term. 〈Limit number: 951. Gap limit: 925.〉
76
Grain Rain, third month, major term. 〈Limit number: 900. Gap limit: 879.〉
77
Summer begins, fourth month, minor term. 〈Limit number: 857. Gap limit: 840.〉
78
滿 [14] [15]
Lesser Fullness, fourth month, major term. 〈Limit number: 823. See note [14]. Gap limit: 812.〉 Editorial note 15.
79
Grain in Ear, fifth month, minor term. 〈Limit number: 800. Gap limit: 799.〉
80
[16]
Summer solstice, fifth month, major term. 〈Limit number: 798. Gap limit: 801.〉 Editorial note 16.
81
Lesser Heat, sixth month, minor term. 〈Limit number: 805. Gap limit: 815.〉
82
Greater Heat, sixth month, major term. 〈Limit number: 825. Gap limit: 842.〉
83
Autumn begins, seventh month, minor term. 〈Limit number: 859. Gap limit: 883.〉
84
End of Heat, seventh month, major term. 〈Limit number: 907. Gap limit: 935.〉
85
White Dew, eighth month, minor term. 〈Limit number: 962. Gap limit: 992.〉
86
Autumn equinox, eighth month, major term. 〈Limit number: 1,021. Gap limit: 1,051.〉
87
Cold Dew, ninth month, minor term. 〈Limit number: 1,080. Gap limit: 1,107.〉
88
[17]
Frost Descends, ninth month, major term. 〈Limit number: 1,133. Gap limit: 1,157.〉 Editorial note 17.
89
[18]
Winter begins, tenth month, minor term. 〈Limit number: 1,181. Per note [18], gap limit: 1,198.〉
90
Lesser Snow, tenth month, major term. 〈Limit number: 1,215. Gap limit: 1,229.〉
91
滿
To find extinction and disappearance days: when the winter-solstice day count carries a fractional remainder, increment the day total by one, multiply by 67,315, divide by 967 for the sexagenary day line, and keep the residue as the fractional part. Reduce the day line modulo sixty, index it against the sexagenary cycle, and the result names the first extinction day after the prior year's winter solstice.
92
滿
For each subsequent extinction day add 69 to the day count and 592 to the fraction, carrying overflow by 967 into the day count, then read off the date as before. When the fractional part reaches zero, the day is a full extinction day rather than a mere disappearance day.
93
The days when Wood, Fire, Metal, and Water begin to govern are found at Establishing Spring, Establishing Summer, Establishing Autumn, and Establishing Winter. Subtract 18 days, 483 parts, and 6 minute-fractions from each of those four dates, then index the result to obtain Earth's governing day immediately before each seasonal establishment. If the day count is too small, add 60 before subtracting. If the fractional part is too small, borrow one day and add 1,843 to the fraction. If the minute-fraction falls short, borrow one fractional unit and add 12 to the small remainder.
94
滿
For hexagram rulership days, take the winter-solstice sexagenary line and multiply its fractional part by six to fix the day when Kan governs. Add 10,091 to the fractional part, carrying overflow by 11,058 into the day count, to reach the day when Zhong Fu governs.
95
Advance each subsequent hexagram by adding 6 to the day count and 967 to the fractional part. For the four cardinal hexagrams, use each one's midpoint day and multiply its fractional part by six.
96
滿 宿滿宿 退
To find the sun's ecliptic longitude, multiply accumulated conjunction days by 1,843, reduce modulo 673,150, then divide by 1,843 for whole degrees and keep the residue as arc-minutes. Index the result from five lodges before Ox through the twenty-eight lodges; the remainder gives the sun's longitude at midnight on the eleventh month's conjunction. For each following day add one degree without adding fractional parts; when crossing the Dipper lodge subtract its special fraction, and if the fraction is too small, step back one degree.
97
滿
To find the moon's longitude at the eleventh month's conjunction midnight, multiply accumulated days by 24,638, reduce modulo 673,150, divide by 1,843 for degrees and fractional parts, and index through the lodges as above.
98
滿 [19]
For a short month add 22 degrees and 806 fractional parts to the moon's position. For a long month add one day, 13 degrees, and 679 fractional parts. Carry every 1,843 fractional parts into one degree to obtain the moon's longitude at the next conjunction midnight. During the last ten days of winter, make a special notation whenever the moon stands in Zhang or Heart. Editorial note 19.
99
滿 滿[20]
To find the longitude of syzygy, multiply the conjunction fractional day by 19; divide by 47 for large fractional parts and keep the remainder as small fractional parts. Add the large fractional parts to the sun's midnight longitude on conjunction day, carrying 1,843 fractional parts per degree; per note [20], index as before to fix the shared solar-lunar longitude for the eleventh month's syzygy.
100
滿滿
For the following month add 29 degrees, 977 large fractional parts, and 42 small fractional parts, propagating carries through the small-fraction, large-fraction, and degree levels. Subtract the Dipper lodge's special fraction when crossing it to obtain the next month's syzygy longitude.
101
滿滿滿
For the sun at first quarter, add 7 degrees, 705 large parts, 10 small parts, and 1 minute-fraction to the syzygy longitude, carrying overflows at each level, then index the result to name first quarter. Repeat the same addition chain to reach full moon, third quarter, and the next syzygy.
102
滿
For the moon at first quarter, add 98 degrees, 1,279 large parts, and 34 small parts to the syzygy longitude, carrying overflows and indexing as before. Repeat the addition to reach full moon, third quarter, and the next syzygy.
103
To find dusk and dawn longitudes, multiply the era divisor for the sun or the lunar circuit for the moon by the nearest term's night water-clock reading, then divide by 200 for the dawn arc. Subtract that dawn arc from the era divisor for the sun or from the lunar circuit for the moon to obtain the dusk arc. Add each arc to the midnight longitude and convert by the appropriate divisor to degrees.
104
滿 滿 [21]
To find nodal distance at syzygy, take the era's accumulated conjunction fraction, add its eclipse-node difference rate, reduce modulo 790,110, and the remainder is the eleventh-month conjunction's distance from the node. Add 134,630 and reduce modulo 790,110 to obtain the next month's nodal distance at conjunction. Add 67,315 to each month's conjunction nodal distance and reduce modulo 790,110 to fix the full-moon nodal distance for that month. When the nodal distance at syzygy or opposition falls between 67,315 and 722,795 inclusive, per note [21], the new moon crosses the node and the full moon brings a lunar eclipse.
105
滿 滿
To determine whether the moon lies north or south of the ecliptic, add the era's lower eclipse-node difference rate to its conjunction fraction, reduce modulo twice 790,110, and compare against 790,110: the era's opening polarity fixes whether the eleventh-month conjunction moon is north or south of the path. When the remainder crosses 790,110, north and south reverse.
106
滿滿滿 [22]
For each following month add 134,630, reduce modulo 790,110, and flip polarity whenever the count crosses the nodal communication constant. If the node is crossed before the eclipse, new moon and full moon share the same side of the ecliptic. If the eclipse comes before the nodal crossing, the full moon lies on the opposite side of the path from the new moon. When the nodal-eclipse value is at or below 67,315, per note [22], the node is crossed before the eclipse. When the value reaches 722,795 or above, the eclipse precedes the nodal crossing. If the node is crossed first and the value lies near the threshold, inspect the previous month as well. If the eclipse comes first and the value lies near the threshold, inspect the following month as well.
107
[23] [24]
To convert nodal distance when the node is crossed first, divide the current nodal distance by 4,559, per note [23], to obtain the distance measured back from the node. [24] When the eclipse precedes the crossing, subtract the nodal distance from 790,110 and divide by 4,559; the quotient is the forward nodal distance, and any remainder stays in degrees and parts. At a nodal distance of fifteen degrees or more, the bodies may cross the node but no eclipse occurs. At ten degrees or less a true eclipse occurs; above ten the obscuration is slight, with only the penumbra touching. Express the depth of obscuration as a fraction of fifteen.
108
西 西
To find where a solar eclipse first bites: on the outer path with node crossed before eclipse, obscuration begins at the southwest limb. If the eclipse precedes the crossing, obscuration begins at the southeast limb. On the inner path with crossing before eclipse, obscuration begins at the northwest limb. If the eclipse precedes the crossing, obscuration begins at the northeast limb. Measure the extent of obscuration as above, using fifteen as the divisor. When the node lies at the center, the eclipse is complete. A lunar eclipse occurs at opposition to the sun, and the corner of first obscuration is the mirror image of the solar case.
109
Lunar speed table: degree, excess-deficit rate, accumulated surplus-deficit, motion parts.
110
[25]
Day 1: 14 degrees. 〈14 fractional parts.〉 Excess rate 26; initial surplus per note [25]; motion parts 280.
111
Day 2: 14 degrees. 〈11 fractional parts.〉 Excess rate 23; accumulated surplus 118,534; motion parts 277.
112
Day 3: 14 degrees. 〈8 fractional parts.〉 Excess rate 20; accumulated surplus 223,391; motion parts 274.
113
[26]
Day 4: 14 degrees. 〈5 fractional parts.〉 Excess rate 17; accumulated surplus 314,571; motion parts 271, per note [26].
114
Day 5: 14 degrees. 〈1 fractional part.〉 Excess rate 13; accumulated surplus 392,074; motion parts 267.
115
Day 6: 13 degrees. 〈14 fractional parts.〉 Excess rate 7; accumulated surplus 451,341; motion parts 261.
116
Day 7: 13 degrees. 〈7 fractional parts.〉 Deficit rate zero; accumulated surplus 483,254; motion parts 254.
117
Day 8: 13 degrees. 〈1 fractional part.〉 Deficit rate 6; accumulated surplus 483,254; motion parts 248.
118
Day 9: 12 degrees. 〈16 fractional parts.〉 Deficit rate 10; accumulated surplus 455,900; motion parts 244.
119
Day 10: 12 degrees. 〈13 fractional parts.〉 Deficit rate 13; accumulated surplus 410,310; motion parts 241.
120
Day 11: 12 degrees. 〈11 fractional parts.〉 Deficit rate 15; accumulated surplus 351,043; motion parts 239.
121
Day 12: 12 degrees. 〈8 fractional parts.〉 Deficit rate 18; accumulated surplus 282,658; motion parts 236.
122
Day 13: 12 degrees. 〈5 fractional parts.〉 Deficit rate 21; accumulated surplus 200,596; motion parts 233.
123
Day 14: 12 degrees. 〈3 fractional parts.〉 Deficit rate 23; accumulated surplus 104,857; motion parts 231.
124
Day 15: 12 degrees. 〈5 fractional parts.〉 Excess rate 21; initial shrinkage; motion parts 233.
125
Day 16: 12 degrees. 〈7 fractional parts.〉 Excess rate 19; accumulated shrinkage 95,739; motion parts 235.
126
Day 17: 12 degrees. 〈9 fractional parts.〉 Excess rate 17; accumulated shrinkage 182,360; motion parts 237.
127
Day 18: 12 degrees. 〈12 fractional parts.〉 Excess rate 14; accumulated shrinkage 259,863; motion parts 240.
128
Day 19: 12 degrees. 〈15 fractional parts.〉 Excess rate 11; accumulated shrinkage 323,689; motion parts 243.
129
Day 20: 12 degrees. 〈18 fractional parts.〉 Excess rate 8; accumulated shrinkage 373,838; motion parts 246.
130
Day 21: 13 degrees. 〈3 fractional parts.〉 Excess rate 4; accumulated shrinkage 410,310; motion parts 250.
131
Day 22: 13 degrees. 〈7 fractional parts.〉 Deficit rate zero; accumulated shrinkage 428,546; motion parts 254.
132
Day 23: 13 degrees. 〈12 fractional parts.〉 Deficit rate 5; accumulated shrinkage 428,546; motion parts 259.
133
Day 24: 13 degrees. 〈18 fractional parts.〉 Deficit rate 11; accumulated shrinkage 405,751; motion parts 265.
134
Day 25: 14 degrees. 〈5 fractional parts.〉 Deficit rate 17; accumulated shrinkage 355,602; motion parts 271.
135
[27]
Day 26: 14 degrees. 〈11 fractional parts.〉 Deficit rate 23; accumulated shrinkage 278,099, per note [27]; motion parts 277.
136
Day 27: 14 degrees. 〈11 fractional parts.〉 Deficit rate 24; accumulated shrinkage 173,242; motion parts 278.
137
Anomalistic week day: 14 degrees. 〈13 fractional parts with micro-fraction 626.〉 Deficit rate 25. 〈Micro-fraction component 626.〉 Accumulated shrinkage 63,826; motion parts 279. 〈Micro-fraction component 226.〉
138
滿
To place a syzygy inside the lunar speed table: add the tabulated anomaly increment to the lunation fraction, reduce modulo 125,621, divide by 4,559 for whole days plus remainder, and index the result to name the anomaly day for the eleventh month's conjunction.
139
[28] 滿滿
For the following month add one day and 4,450 day-fraction parts. Per note [28], for opposition add fourteen days and 3,489 fractional parts. Carry fractions into days at 4,559, then reduce full weeks of 27 days. If the fractional division underflows at the week boundary, borrow one day and add the weekly complement 2,031.
140
[29] 滿 [30]
Per note [29], to fix syzygy instants multiply the anomalistic day fraction by the tabulated rate and apply it to the surplus-deficit column for the corrected integral. Divide the adjusted integral by the difference between 19 and the tabulated motion parts, then subtract surplus or add shrinkage to the base fractional day. A carry past the day divisor pushes the conjunction instant into the next civil day. A borrow moves the corrected instant onto the previous day. Lunar eclipses take their clock time directly from the corrected remainder line. At the anomalistic week boundary, multiply the deficit column by the weekly fractional day for the corrected integral. Per note [30], combine the rate product with weekly micro-parts to obtain the post-boundary correction stack. Finish the week-crossing correction by dividing with the prescribed divisor mix, then add to the base fraction as above.
141
滿 滿
Multiply the corrected fraction by twelve and divide by 4,559 to name the twelve double-hours from midnight. Split the leftover into quarters of the divisor for the shao, ban, and tai fine subdivisions. Further triple the tail to reach the strong step, rounding up at half-divisor. Add strong units to weak, half, or full subdivisions per the classical clepsydra notation. Two strongs collapse into a weak grade, stepping through the ladder to a full weak double-hour mark. Read the final label against the stem hour to recover shao, tai, ban, qiang, and ruo. When an eclipse full moon lies within four days of a major term, consult the tight limit number. Beyond five days from the major term, apply the wider gap-limit rule. When the corrected fraction falls under both margin thresholds, promote the count to the next day.
142
Dipper: 26 degrees. 〈455 fractional parts.〉 Ox 8°, Maiden 12°, Emptiness 10°, Rooftop 17°, House 16°, Wall 9°.
143
Northern quadrant: 98°. 〈455 fractional parts.〉
144
Straddles 16°, Harvest 12°, Stomach 14°, Hairy Head 11°, Net 16°, Turtle Beak 2°, Triaster 9°.
145
西
Western quadrant: 80°.
146
Well 33°, Ghost 4°, Willow 15°, Star 7°, Extended Net 18°, Wings 18°, Chariot Platform 17°.
147
Southern quadrant: 112°.
148
Horn 12°, Gullet 9°, Base 15°, Chamber 5°, Heart 5°, Tail 18°, Winnowing Basket 11°.
149
Eastern quadrant: 75°.
150
Table columns: major solar term; solar lodge longitude; solar declination from the ecliptic pole; noon shadow length; day clepsydra marks; night clepsydra marks; dusk culmination; dawn culmination.
151
Winter solstice. 〈Eleventh month, major term.〉 Dipper 21. 〈Shao.〉 Declination 115°; shadow 1 zhang 3 chi; day clepsydra 45; night clepsydra 55; dusk star Straddles 6. 〈Weak.〉 Dawn star Gullet 2. 〈Shao-strong.〉
152
Minor Cold. 〈Twelfth month, minor term.〉 Maiden 2. 〈Shao.〉 Declination 113°. 〈Strong.〉 Shadow 1 zhang 2 chi 3 cun; day clepsydra 45. 〈8 fractional parts.〉 Night clepsydra 54. 〈2 fractional parts.〉 Dusk star Harvest 6. 〈Half-strong.〉
153
Dawn star Base 7. 〈Strong.〉
154
Major Cold. 〈Twelfth month, major term.〉 Emptiness 5. 〈Half-weak.〉
155
[33]
Declination 110°. 〈Very weak.〉 Shadow 1 zhang 1 chi; day clepsydra 46. 〈8 fractional parts.〉 Night clepsydra 53. 〈2 fractional parts.〉 Dusk star Stomach 11. 〈Very strong.〉 Editorial note 33: Heart. 〈Half.〉
156
Establishment of Spring. 〈First month, minor term.〉 Rooftop 10. 〈Very weak.〉 Declination 106°. 〈Shao-weak.〉 Shadow 9 chi 6 cun; day clepsydra 48. 〈6 fractional parts.〉 Night clepsydra 51. 〈4 fractional parts.〉 Dusk star Net 5. 〈Shao-weak.〉 Dawn star Tail 7. 〈Half-weak.〉
157
[34]
Rain Water. 〈First month, major term.〉 House 8. 〈Very strong.〉 Declination 101°. 〈Strong.〉 Shadow 7 chi 9 cun. 〈5 fractional parts.〉 Day clepsydra 50. 〈8 fractional parts.〉 Night clepsydra 49. 〈2 fractional parts.〉 Dusk star Triaster 6. 〈Half-weak.〉 Dawn star Winnowing Basket. 〈Half-weak.〉 Editorial note 34.
158
Awakening of Insects. 〈Second month, minor term.〉 Wall 8. 〈Strong.〉 Declination 95°. 〈Strong.〉 Shadow 6 chi 5 cun; day clepsydra 53. 〈3 fractional parts.〉 Night clepsydra 46. 〈7 fractional parts.〉 Dusk star Well 17. 〈Shao-weak.〉 Dawn star Dipper beginning. 〈Shao.〉
159
Spring equinox. 〈Second month, major term.〉 Straddles 14. 〈Shao-strong.〉 Declination 89°. 〈Shao-strong.〉 Shadow 5 chi 2 cun. 〈5 fractional parts.〉 Day clepsydra 55. 〈8 fractional parts.〉 Night clepsydra 44. 〈2 fractional parts.〉 Dusk star Ghost 4; dawn star Dipper 11. 〈Weak.〉
160
Pure Brightness. 〈Third month, minor term.〉 Stomach 1. 〈Half.〉 Declination 83°. 〈Shao-weak.〉 Shadow 4 chi 1 cun. 〈5 fractional parts.〉 Day clepsydra 58. 〈3 fractional parts.〉 Night clepsydra 41. 〈7 fractional parts.〉 Dusk star Star 4. 〈Very.〉 Dawn star Dipper 21. 〈Half.〉
161
Grain Rain. 〈Third month, major term.〉 Hairy Head 2. 〈Very.〉 Declination 77°. 〈Very strong.〉 Shadow 3 chi 2 cun; day clepsydra 60. 〈5 fractional parts.〉 Night clepsydra 39. 〈5 fractional parts.〉 Dusk star Extended Net 17; dawn star Ox 6. 〈Half.〉
162
[35]
Start of Summer. 〈Fourth month, minor term.〉 Net 6. 〈Very.〉 Editorial note 35: declination 73°. 〈Shao-weak.〉 Shadow 2 chi 5 cun. 〈2 fractional parts.〉 Day clepsydra 62. 〈4 fractional parts.〉 Night clepsydra 37. 〈6 fractional parts.〉 Dusk star Wings 17. 〈Very.〉 Dawn star Maiden 10. 〈Shao-weak.〉
163
滿
Lesser Fullness. 〈Fourth month, major term.〉 Triaster 4. 〈Shao-weak.〉 Declination 69°. 〈Very.〉 Shadow 1 chi 9 cun. 〈8 fractional parts.〉 Day clepsydra 63. 〈9 fractional parts.〉 Night clepsydra 36. 〈1 fractional part.〉 Dusk star Horn. 〈Very weak.〉 Dawn star Rooftop. 〈Very weak.〉
164
Grain in Ear. 〈Fifth month, minor term.〉 Well 10. 〈Half-weak.〉 Declination 67°. 〈Shao-weak.〉 Shadow 1 chi 6 cun. 〈8 fractional parts.〉 Day clepsydra 64. 〈9 fractional parts.〉 Night clepsydra 35. 〈1 fractional part.〉 Dusk star Gullet 5. 〈Very.〉 Dawn star Rooftop 14. 〈Strong.〉
165
Summer solstice. 〈Fifth month, major term.〉 Well 25. 〈Half-strong.〉 Declination 67°. 〈Strong.〉 Shadow 1 chi 5 cun; day clepsydra 65; night clepsydra 35; dawn star Base 12. 〈Shao-weak.〉 Dusk star House 12. 〈Strong.〉
166
Lesser Heat. 〈Sixth month, minor term.〉 Willow 3. 〈Very strong.〉 Declination 67°. 〈Very strong.〉 Shadow 1 chi 7 cun; day clepsydra 64. 〈7 fractional parts.〉 Night clepsydra 35. 〈3 fractional parts.〉 Dusk star Tail 1. 〈Very strong.〉 Dawn star Straddles 2. 〈Very strong.〉
167
[36]
Greater Heat. 〈Sixth month, major term.〉 Star 4. 〈Strong.〉 Declination 70°; shadow 2 chi; day clepsydra 63. 〈8 fractional parts.〉 Night clepsydra 36. 〈2 fractional parts.〉 Dusk star Tail 15. 〈Half-strong.〉 Editorial note 36: Harvest 3. 〈Very.〉
168
Start of Autumn. 〈Seventh month, minor term.〉 Extended Net 12. 〈Shao.〉 Declination 73°. 〈Half-strong.〉 Shadow 2 chi 5 cun. 〈5 fractional parts.〉 Day clepsydra 62. 〈3 fractional parts.〉 Night clepsydra 37. 〈7 fractional parts.〉 Dusk star Winnowing Basket 9. 〈Tai-strong.〉 Dawn star Stomach 9. 〈Tai-weak.〉
169
End of Heat. 〈Seventh month, major term.〉 Solar longitude Wings 9. 〈Half.〉 Declination 78°. 〈Half-strong.〉 Shadow 3 chi 3 cun. 〈3 fractional parts.〉 Day clepsydra 60. 〈2 fractional parts.〉 Night clepsydra 39. 〈8 fractional parts.〉 Dusk star Dipper 10. 〈Shao.〉 Dawn star Net 3. 〈Tai.〉
170
White Dew. 〈Eighth month, minor term.〉 Solar longitude Chariot Platform 6. 〈Tai.〉 Declination 84°. 〈Shao-strong.〉 Shadow 4 chi 3 cun. 〈5 fractional parts.〉 Day clepsydra 57. 〈8 fractional parts.〉 Night clepsydra 42. 〈2 fractional parts.〉 Dusk star Dipper 21. 〈Strong.〉 Dawn star Triaster 5. 〈Shao-strong.〉
171
[37]
Autumn equinox. 〈Eighth month, major term.〉 Solar longitude Horn 5. 〈Weak.〉 Declination 90°. 〈Half-strong.〉 Shadow 5 chi 5 cun, per note [37]; night clepsydra 55. 〈2 fractional parts.〉 Day clepsydra 44. 〈8 fractional parts.〉 Dusk star Ox 5. 〈Shao.〉 Dawn star Well 16. 〈Shao-strong.〉
172
[38]
Cold Dew. 〈Ninth month, minor term.〉 Solar longitude Gullet 8. 〈Half-weak.〉 Per note [38], declination 96°. 〈Tai-strong.〉 Shadow 6 chi 8 cun. 〈5 fractional parts.〉 Day clepsydra 52. 〈6 fractional parts.〉 Night clepsydra 47. 〈4 fractional parts.〉 Dusk star Maiden 7. 〈Tai.〉 Dawn star Ghost 3. 〈Shao-strong.〉
173
Frost Descends. 〈Ninth month, major term.〉 Solar longitude Base 14. 〈Shao-strong.〉 Declination 102°. 〈Shao-strong.〉 Shadow 8 chi 4 cun; day clepsydra 50. 〈3 fractional parts.〉 Night clepsydra 49. 〈7 fractional parts.〉 Dusk star Emptiness 6. 〈Tai.〉 Dawn star Star 3. 〈Tai.〉
174
[39]
Winter begins. 〈Tenth month, minor term.〉 Solar longitude Tail 4. 〈Half-strong.〉 Declination 107°. 〈Shao-strong.〉 Shadow 1 zhang, per note [39]; day clepsydra 48. 〈2 fractional parts.〉 Night clepsydra 51. 〈8 fractional parts.〉 Dusk star Rooftop 8. 〈Strong.〉 Dawn star Extended Net 15. 〈Tai-strong.〉
175
[40] [41]
Lesser Snow. 〈Tenth month, major term.〉 Solar longitude Winnowing Basket 1. 〈Tai-strong.〉 Declination 111°. 〈Weak.〉 Shadow 1 zhang 1 chi 4 cun; night clepsydra 46. 〈7 fractional parts.〉 Day clepsydra 53. 〈3 fractional parts.〉 Dusk star House 3. 〈Half-strong.〉 Per note [40], dawn star Wings 15. 〈Tai.〉 Editorial note 41.
176
[42] [43]
Greater Snow. 〈Eleventh month, minor term.〉 Solar longitude Dipper 6; declination 113°. 〈Tai-strong.〉 Shadow 1 zhang 2 chi 5 cun. 〈6 fractional parts.〉 Day clepsydra 45. 〈5 fractional parts.〉 Per note [42], night clepsydra 54. 〈5 fractional parts.〉 Dusk star Wall. 〈Half-strong.〉 Dawn star Chariot Platform 15. 〈Shao-strong.〉 Editorial note 43.
177
[44]
The median terms listed at right are found by the procedure above; the first result is the winter solstice in the middle of the eleventh month. Adding one step yields the next minor term; adding a minor term in turn yields the corresponding major term. Culmination stars are reckoned from the sun's lodge position on the given day. For the target year, take each solar term's fractional remainder, multiply by four and divide by the day divisor for shao units; triple the leftover and divide again for qiang units. Subtract those shao and qiang amounts from the tabulated dusk and dawn culmination stars to obtain the corrected values. Editorial note 44.
178
Procedure for the five planets:
179
宿 [45] [46] [47] [48]
The five planets are Jupiter (Year Star), Mars (Sparkling Deluder), Saturn (Reimbursing Star), Venus (Great White), and Mercury (Chronogram Star). Each planet moves now slowly, now swiftly, sometimes halting and sometimes moving backward. Since the world's beginning, when yin and yang first divided, the sun, moon, and five planets all clustered in the Star Chronogram lodge. Starting from Star Chronogram they travel the sky in concert, their slow, fast, stationary, and retrograde motions continually catching and passing one another. When a planet and the sun share the same lodge and degree, the event is termed a conjunction. The interval from one conjunction day to the next is one complete planetary cycle. Reduce each planet's cycle length and the civil year to a common fractional base; the resulting year ratio is the total cycle year-count and the conjunction ratio is the total cycle conjunction-count. Editorial note 45. With both rates established, the computational divisors follow. Form the conjunction-month divisor by rule years times conjunction-count, the day-degree divisor by era divisor times conjunction-count, and conjunction-month parts by rule months times year-count; divide for whole conjunction months and keep the remainder as month remainder. Multiply conjunction months by the communication number, divide by the day divisor for the sexagenary line, reduce modulo sixty, and the result is the star-conjunction new-moon large remainder. What remains after the sexagenary reduction is the new-moon small remainder. Per note [46], combine month remainder and new-moon fraction with the prescribed divisors; the quotient is the day-count on which the planetary conjunction enters the lunar month. Reduce the leftover by the communication divisor; per note [47], the result is the fractional day remainder within the month. Per note [48], subtract the new-moon small remainder from the day divisor; the difference is the new-moon void fraction. Multiply the calendar's dipper fraction by the conjunction-count to obtain the planetary dipper fraction. For Jupiter, Mars, and Saturn, subtract the conjunction-count from the year-count, multiply by the circuit-of-heaven constant, divide by the day-degree divisor for whole degrees, and keep the remainder as degree remainder. For Venus and Mercury, multiply the year-count by the circuit-of-heaven constant and divide by the day-degree divisor for degrees and degree remainder.
180
Jupiter: total cycle year-count, 1,255.
181
Total cycle conjunction-count, 1,149.
182
Conjunction-month divisor, 21,831.
183
Day-degree divisor, 2,117,607.
184
Conjunction-month count, 13.
185
Month remainder, 11,122.
186
New-moon large remainder, 23.
187
New-moon small remainder, 4,093.
188
Day entered in month, 15.
189
Day remainder, 1,995,664.
190
New-moon void fraction, 466.
191
Dipper fraction, 522,795.
192
Planetary degree, 33.
193
[49]
Degree remainder, 1,472,869. Editorial note 49.
194
Mars: total cycle year-count, 5,105.
195
Total cycle conjunction-count, 2,388.
196
Conjunction-month divisor, 45,372.
197
Day-degree divisor, 4,401,084.
198
Conjunction-month count, 26.
199
Month remainder, 20,003.
200
New-moon large remainder, 47.
201
New-moon small remainder, 3,627.
202
Day entered in month, 13.
203
Day remainder, 3,585,230.
204
New-moon void fraction, 932.
205
Dipper fraction, 1,086,540.
206
Planetary degree, 50.
207
Degree remainder, 1,412,150.
208
Saturn: total cycle year-count, 3,943.
209
Total cycle conjunction-count, 3,809.
210
Conjunction-month divisor, 72,371.
211
Day-degree divisor, 7,019,987.
212
Conjunction-month count, 12.
213
Month remainder, 58,153.
214
New-moon large remainder, 54.
215
New-moon small remainder, 1,674.
216
Day entered in month, 24.
217
Day remainder, 675,364.
218
New-moon void fraction, 2,885.
219
Dipper fraction, 1,733,095.
220
Planetary degree, 12.
221
Degree remainder, 5,962,256.
222
Venus: total cycle year-count, 1,907.
223
Total cycle conjunction-count, 2,385.
224
Conjunction-month divisor, 45,315.
225
Day-degree divisor, 4,395,555.
226
Conjunction-month count, 9.
227
Month remainder, 40,310.
228
New-moon large remainder, 25.
229
New-moon small remainder, 3,535.
230
Day entered in month, 27.
231
Day remainder, 194,990.
232
New-moon void fraction, 1,024.
233
Dipper fraction, 1,851,175.
234
Planetary degree, 292.
235
Degree remainder, 194,990.
236
Mercury: total cycle year-count, 1,870.
237
Total cycle conjunction-count, 11,789.
238
Conjunction-month divisor, 223,991.
239
Day-degree divisor, 21,721,727.
240
Conjunction-month count, 1.
241
Month remainder, 215,459.
242
New-moon large remainder, 29.
243
New-moon small remainder, 2,419.
244
Day entered in month, 28.
245
[50]
Day remainder, 20,344,261. Editorial note 50.
246
New-moon void fraction, 2,140.
247
Dipper fraction, 5,363,995.
248
Planetary degree, 57.
249
Degree remainder, 20,344,261.
250
滿
The procedure for the five planets runs thus: take the years since the Renchen origin, multiply by the total cycle conjunction-count, and divide by the total cycle year-count; the quotient is the accumulated conjunction, the remainder the conjunction surplus. Subtract the conjunction remainder from the total cycle conjunction-count: a result of one places the conjunction in the prior year, of two in the year before that; if nothing remains, the conjunction falls in the year sought. Subtract that remainder from the total cycle conjunction-count to obtain the degree fraction. For Venus and Mercury, an even accumulated conjunction marks a morning appearance, an odd one an evening appearance.
251
滿 [51]滿
To find the conjunction month, multiply both the month count and month remainder by the accumulated conjunction; carry into whole months whenever the remainder fills the conjunction-month divisor, and keep the leftover as month remainder. Divide accumulated months by the era month-total; the quotient names the era entered and the remainder the month within that era. Then multiply the auxiliary by rule intercalations; per note [51], divide by rule months for whole intercalary months, subtract them from the era month, remove full years from what remains, and index from the celestial first month—the result is the star-conjunction month. If the date falls at an intercalary boundary, govern it by the nearest new moon.
252
滿
To obtain the conjunction new moon, multiply the era month by the communication number, divide by the day divisor for accumulated days, and keep the remainder as the small fraction. Reduce accumulated days modulo sixty for the sexagenary remainder, index from the era entered, and the result is the star-conjunction new-moon day.
253
滿滿
For the day within the month, combine month remainder and new-moon fraction through the communication and conjunction divisors; divide by the day-degree divisor for the entry day, keeping any remainder as day surplus. Index that day from the new moon outside the reckoning to name the day entered in the month.
254
滿
For the conjunction degree, multiply the circuit of heaven by the degree fraction and divide by the day-degree divisor; index from five degrees before Ox to obtain the lodge and degree of conjunction.
255
滿滿 滿 滿
To find the next conjunction month, add the tabulated month count and remainder to the current year-month; carry a month whenever the remainder fills the conjunction-month divisor—if the month still lies within the year, the conjunction falls in that year; otherwise remove full years, accounting for intercalation, and the remainder points to a later year; if it fills once more, the conjunction lies two years ahead. For Venus and Mercury, adding to a morning date yields an evening one, and adding to an evening date yields a morning one.
256
[52]滿
For the next conjunction new moon, add the tabulated new-moon remainders to those of the conjunction month, and when the month fraction carries a month, also add twenty-nine to the large remainder and 2,419 to the small. Per note [52], carry from small to large remainder when the small fraction fills the day divisor, then index as before.
257
[53][54]滿 滿 滿 滿 宿
Per note [53], add the tabulated entry day and day remainder to the current ones; per note [54], carry a day whenever the remainder fills the day-degree divisor. If the preceding conjunction new-moon small remainder fills its void fraction, subtract one day; if the later small remainder reaches 2,419 or above, subtract twenty-nine days; otherwise subtract thirty days; what remains is the next conjunction day within the month, indexed from the new moon. For the next conjunction degree, add the tabulated degree and fraction and index the lodge sequence as before.
258
退 西
Jupiter: conjoined with the sun at dawn, it disappears; moving prograde for sixteen days plus 997,832 fractional parts and two degrees plus 1,795,238 fractional parts, it reappears in the east at dawn, behind the sun. Prograde and swift: it advances eleven parts per fifty-seven each day, eleven degrees in fifty-seven days. Prograde and slow: nine parts per day, nine degrees in fifty-seven days, then it stops. It remains motionless for twenty-seven days, then reverses. Retrograde: one part in seven per day, twelve degrees back in eighty-four days, then another twenty-seven-day halt. Slow again at nine parts per day, nine degrees in fifty-seven days, then prograde once more. Swift motion at eleven parts per day, eleven degrees in fifty-seven days; standing ahead of the sun, it sets in the west at dusk. Prograde for sixteen days plus 997,832 fractional parts and two degrees plus 1,795,238 fractional parts, then it conjoins with the sun. One complete cycle: 398 days plus 1,995,664 fractional parts; planetary travel 33 degrees plus 1,472,869 fractional parts.
259
退 西
Mars: conjoined with the sun at dawn, it disappears; after seventy-two days plus 1,792,615 fractional parts and fifty-six degrees plus 1,249,345 fractional parts, it reappears in the east at dawn, behind the sun. Prograde: fourteen parts in twenty-three per day, one hundred twelve degrees in one hundred eighty-four days. Prograde and slow again: twelve parts per day, forty-eight degrees in ninety-two days, then it stops. Motionless for eleven days, then it reverses. Retrograde: seventeen parts in sixty-two per day, seventeen degrees back in sixty-two days, then another eleven-day halt. Prograde and slow again: twelve parts per day, forty-eight degrees in ninety-two days, then swift motion resumes. Fourteen parts per day, one hundred twelve degrees in one hundred eighty-four days; ahead of the sun, it sets in the west at dusk. Prograde for seventy-two days plus 1,792,615 fractional parts and fifty-six degrees plus 1,249,345 fractional parts, then it conjoins with the sun. One complete cycle: 780 days plus 3,585,230 fractional parts; planetary travel 415 degrees plus 2,498,690 fractional parts.
260
退 西
Saturn: conjoined with the sun at dawn, it disappears; after nineteen days plus 3,847,675½ fractional parts and two degrees plus 6,491,121½ fractional parts, it reappears in the east at dawn, behind the sun. Prograde: thirteen parts in 172 per day, six and a half degrees in eighty-six days, then it stops. Motionless for thirty-two and a half days, then it reverses. Retrograde: one part in seventeen per day, six degrees back in one hundred two days, then another halt. After thirty-two and a half motionless days it turns prograde again at thirteen parts per day, six and a half degrees in eighty-six days; ahead of the sun, it sets in the west at dusk. Prograde for nineteen days plus 3,847,675½ fractional parts and two degrees plus 6,491,121½ fractional parts, then it conjoins with the sun. One complete cycle: 378 days plus 675,364 fractional parts; planetary travel 12 degrees plus 5,962,256 fractional parts.
261
退 退
Venus: conjoined with the sun at dawn, it disappears; after six days retreating four degrees it reappears in the east at dawn, behind the sun and moving backward. Slow and retrograde: three parts in five per day, six degrees back in ten days. Motionless for seven days, then it reverses. Prograde and slow: thirty-three parts in forty-five per day, thirty-three degrees in forty-five days, then faster prograde. Swift: one degree plus fourteen parts in ninety-one per day, one hundred five degrees in ninety-one days, still prograde. Still faster: one degree plus twenty-one parts in ninety-one per day, one hundred twelve degrees in ninety-one days; behind the sun, it vanishes in the east at dawn. Prograde for forty-two days plus 194,990 fractional parts and fifty-two degrees plus 194,990 fractional parts, then it conjoins with the sun. One conjunction: 292 days plus 194,990 fractional parts; planetary travel the same.
262
西 退西 退
Venus: conjoined with the sun at dusk, it disappears; prograde for forty-two days plus 194,990 fractional parts and fifty-two degrees plus 194,990 fractional parts, then reappears in the west at dusk, ahead of the sun. Prograde and swift: one degree plus twenty-one parts in ninety-one per day, one hundred twelve degrees in ninety-one days, then slower prograde. Slow: one degree plus fourteen parts per day, one hundred five degrees in ninety-one days, still prograde. Still slower: thirty-three parts in forty-five per day, thirty-three degrees in forty-five days, then it stops. Motionless for seven days, then it reverses. Retrograde: three parts in five per day, six degrees back in ten days; ahead of the sun, it sets in the west at dusk. Retrograde for six days, four degrees back, then it conjoins with the sun. Two conjunctions complete one cycle: 584 days plus 389,980 fractional parts; planetary travel the same.
263
退 退
Mercury: conjoined with the sun at dawn, it disappears; after eleven days retreating seven degrees it reappears in the east at dawn, behind the sun. Retrograde and swift: one degree back in one day, then it stops. Motionless for one day, then it reverses. Prograde and slow: seven parts in eight per day, seven degrees in eight days, then faster prograde. Swift: one degree plus four parts in eighteen per day, twenty-two degrees in eighteen days; behind the sun, it vanishes in the east at dawn. Prograde for eighteen days plus 20,344,261 fractional parts and thirty-six degrees plus 20,344,261 fractional parts, then it conjoins with the sun. One conjunction: 57 days plus 20,344,261 fractional parts; planetary travel the same.
264
西 退西 退
Mercury: conjoined with the sun at dusk, it disappears; after eighteen days plus 20,344,261 fractional parts and thirty-six degrees plus 20,344,261 fractional parts it reappears in the west at dusk, ahead of the sun. Prograde and swift: one degree plus four parts in eighteen per day, twenty-two degrees in eighteen days, then slower prograde. Slow: seven parts in eight per day, seven degrees in eight days, then it stops. Motionless for one day, then it reverses. Retrograde: one degree back in one day; ahead of the sun, it sets in the west at dusk. Retrograde for eleven days, seven degrees back, then it conjoins with the sun. Two conjunctions complete one cycle: 115 days plus 18,961,395 fractional parts; planetary travel the same.
265
滿 滿 [55]
Five-star step procedure: add the tabulated hidden-day and conjunction degree remainders; carry a whole unit when the sum fills the day-degree divisor, then index as before for the star's appearance day and fractional degree. Multiply the appearance degree fraction by the motion denominator, divide by the day-degree divisor rounding up at half, then add the daily motion fraction until a full degree is reached. When retrograde and prograde use different denominators, convert the prior fraction by the ratio of denominators to obtain the motion fraction for the current phase. Stationary phases carry forward the prior value, retrograde phases subtract, hidden phases omit degrees, dipper fractions are cleared; per note [55], use the motion denominator as the rate. When fractional parts increase or decrease, successive phases adjust one against the other.
266
退
Though the five planets generally follow broad patterns of slow, swift, stationary, and retrograde motion, their trespasses, halts, and reversals cannot always be forced by computation alone. If even the moon shows slow and fast intervals in its course, the five planets do so all the more. Only the sun moves through Heaven at a fixed pace, advancing and retreating by a steady rate—neither lagging nor rushing, neither straying outward nor inward. That is the virtue of the sovereign.
267
To derive Jupiter's total cycle year-count, multiply its day-degree divisor by one complete cycle in days and fractional parts, then divide by the circuit of heaven.
268
滿
For Jupiter's total cycle conjunction-count, multiply the day-degree divisor by the circuit of heaven, reduce by the era divisor, then divide again by the circuit of heaven. The same method applies to all five planets.
269
In the first year of Huangchu of Wei, the eleventh month was short; the obscuration commenced on Jimao in a Jihai year; on the eleventh month, Jimao day, at new moon and winter solstice, Wei submitted this memorial."
270
Shu under the Liu clan never adopted a new calendar and presumably kept the Han Quarter-Remainder system. Kan Ze, Director of the Secretariat in Wu, learned Liu Hong's Supernal Icon calendar from Xu Yue of Donglai, known as Gonghe. The Sun regime therefore employed the Supernal Icon calendar until Wu fell.
271
使
The historiographer observes that in Zou Yan's scheme of the Five Phases, Zhou belonged to Fire. Zou Yan lived during the Zhou age; he can hardly have been unaware of Zhou's assigned phase. Moreover, the Zhou calendar ran eight hundred years, and the Qin house already held the record of Zhou's founding. The question of Zhou as Fire or Wood is easily settled on the evidence. On the rotation of dynastic phases, only two major schools of interpretation exist. Zou Yan built his theory on mutual conquest; Liu Xiang on mutual generation. On these grounds, no serious account can escape one of the two schools. Suppose one accepts Liu Xiang's claim that Zhou was Wood and that Qin, replacing Zhou, altered the phase succession. On the conquest theory, Wood should yield to Metal; on the generation theory, Wood ought to produce Fire. Yet Qin proclaimed itself Water—a claim that fits neither scheme cleanly. Liu Xiang's argument therefore fails to convince. Although Zhang Cang was a Han official, he had lived into Zhou times, served Qin as Keeper of the Pillars below, and had access to the full archive. Qin suppressed the classics but not calendrical science; enough Zhou material survived to show that Han's claim to Water was not fabricated from nothing. Jia Yi cites Qin as saying, "Han belongs to Earth." —meaning that Han supplanted Qin. Each of the two theories, examined closely, has its own internal logic. Zhang Cang argued that Han Water conquered Zhou Fire and therefore excluded Qin from the phase sequence. Jia Yi held that Han Earth conquered Qin Water and counted Qin as a legitimate dynasty. However they differ on Qin and Han, both agree that Zhou was Fire. On balance, the conquest theory fits the historical record better. Following Zhang Cang and demoting Qin yields Han Water, Wei Earth, Jin Wood, and Song Metal; following Jia Yi and retaining Qin yields Han Earth, Wei Wood, Jin Metal, and Song Fire. One critic asks: "When Han Gaozu cut the serpent, the divine mother wept in the night that the Red Emperor's son had slain the White Emperor's son—does this not prove Han was Fire?" That objection, too, misses the point. If Han were truly Fire, the omen should have named the Red Emperor himself, not his son. And what, in any case, is meant by the White Emperor's son? The answer is that Han was Earth—Earth born of Fire—and Qin was Water—Water born of Metal. Han as Earth was therefore the Red Emperor's son, and Qin as Water the White Emperor's son. The critic adds: "Liu Xiang taught that the Five Phases mutually conquer; how then can Earth also be the Red Emperor's son?" The reply is that the Five Phases admit both a logic of conquest and a logic of generation. Neither principle cancels the other. In conquest, Earth overcomes Water; in generation, Earth is born of Fire—the two relationships are simply different."
272
[56] [57]
Cui Shi's Monthly Ordinances for the Four Peoples states: "Zu is the spirit of the road." The Yellow Emperor's son Leizu loved distant journeys and died on the road; for that reason he is honored as the road spirit. Ji Han's Preface to the Fu on the Road Spirit records: [56] Han used Bingwu, Wei Dingwei, and Jin the you day of the first month. No one can say with certainty when the cult of Zu began. Some hold that travelers pray to the road spirit and call the rite Zu: a gentleman departing on duty sets out the offering along the route; a family about to move a coffin invokes the name on the courtyard steps. Others say that after many generations names and tombs are lost and spirits have no temple to receive them; at the year's first auspicious day people therefore raise decorated canopies and bright banners to summon the dead, that all the ancestors may draw near. Editorial note 57.
273
[58]
Under Emperor Wu of Jin, Palace Attendant Liu Zhi of Pingyuan [58] proposed a three-hundred-year Dipper-calendar reform, arguing that the Quarter-Remainder system drops one day every three centuries, with 150 as degree divisor and 37 as dipper fraction. He padded the proposal with flattering rhetoric to bolster his case. Wang Shuo, Central Commander of the Left in the Eastern Jin, admired the method because its upper origin fell in a jiazi year and wanted to treat a jiazi ninety-seven thousand years from creation as the world's beginning—what He Chengtian dismissed as a flawed foundation. The Jingchu calendar's noon shadow measurements relied on the Han Quarter-Remainder method and therefore drifted steadily out of true. Its planetary calculations were especially coarse. After the Eastern Jin the Supernal Icon planetary method replaced it, yet still showed forward and backward error.
274
Emperor Wu of Song took a keen interest in calendrical science, and the Crown Prince's Director of Standards He Chengtian privately composed a new system. In the twentieth year of Yuanjia he submitted a memorial stating:
275
Your subject is dull by nature and understood little in his youth. From boyhood I have loved calendrical work and have pursued it with single-minded devotion into old age. My late uncle, former Director of the Secretariat Xu Guang, was expert in these matters and kept a running ledger of the seven-luminaries calendar, noting every hit and miss. From the Taihe era to the end of Taiyuan spans roughly forty years. I have tested and corrected that record in recent years, adding another forty years to the present. Its errors and accuracies can therefore be known in detail.
276
Heaven's pivot turns without rest and the seven luminaries run appointed courses; though their meetings and partings follow broad regularities, the interplay of old and new cycles produces tiny discrepancies that, day by day and year by year, accumulate into plain error. Hence the Yu Documents teaches reverent accord with Heaven, and the Changes instructs us to regulate the calendar—meaning that the calendar should follow Heaven, not force Heaven to follow the calendar. Han observers mixed measurements at the Clear Terrace, using dusk, dawn, and meridian stars to test the sun's place; though the sun itself is invisible, a full moon must eclipse at opposition, and from the moon one can infer the sun's lodge. To leave the easy path and exhaust oneself over the difficult—this your subject cannot fathom.
277
[59]
The Canon of Yao says, "The day is longest and the star is Fire—thereby fix midsummer." Today, at the last month of summer, Fire crosses the meridian. It also says, "At midnight the star is Emptiness—thereby mark mid-autumn." Today, at the last month of autumn, Emptiness crosses the meridian. Across the twenty-seven centuries since, meridian-star tests show a drift of twenty-seven or twenty-eight degrees. At the time of Yao's ordinance, winter solstice would have placed the sun near ten degrees of Maiden. The Han Taichu calendar placed winter solstice at the start of Ox; the Later Han Quarter-Remainder and Wei Jingchu systems both placed it at Dipper 21. Testing by lunar eclipse, I find that Jingchu's winter solstice today should fall at Dipper 17. Historiographical officers, under edict, measured solstice shadows with the earth-standard and found a discrepancy of more than three days. Cross-checking accumulated years and reports from Jiaozhou confirms the same drift. Today's solstices are therefore not Heaven's true solstices. Heaven's southern extreme: [59] the sun now stands at Dipper 13 or 14. This implies that the nineteen-year, seven-intercalation rule runs slightly high. Constantly rewriting rules only multiplies computation; the calendar should be revised when needed to recover agreement with Heaven. The Later Han Treatise notes that the spring equinox day is longer and the autumn equinox day shorter, by more than half a quarter. Since the equinoxes fall between the solstices yet differ in length, one sees that the spring equinox lies nearer summer and therefore yields a longer day; while the autumn equinox lies nearer winter and therefore yields a shorter one. Yang Wei failed to grasp this and applied the values anyway; his memorial on the superior calendar boasts, "From antiquity to the present no calendar has matched its own perfection." Yet how could he fail to see the problem and still make such a claim? Your subject therefore proposes the Yuanjia calendar: era length 608, degree divisor 304, lodge fraction 75, year beginning at the month of Establishment of Spring, first qi at Rain Water, and era heads in years whose intercalary remainder equals one. Winter solstice is placed at the third day, fifth watch, before the upper day. The sun's position is moved four degrees from the old placement. The moon, moreover, has slow and fast intervals; syzygy and eclipse need not fall exactly on new or full moon—another point the old methods mishandle. The Yuanjia system therefore uses surplus-deficit correction on the small remainder to fix the days of new and full moon.
278
Your Majesty, extending the Way of the sages and anticipating Heaven without offense, labors over the realm's affairs and illuminates the great enterprise; probing former texts and seeking principles not yet understood, you exhaust spirit and comprehend change—nothing lies beyond your view. Your humble subject, meeting such enlightened times, offers what little a narrow tube or needle's eye can contribute. I respectfully ask that the Yuanjia method I submit be referred to the historiographical officers for examination of its accuracy. If any part proves usable despite its flaws, it may yet help correct errors and fill gaps.
279
Edict: "He Chengtian's presentation is exceptionally well founded. Send it out for detailed review."
280
Grand Astrologer Qian Lezhi and Assistant Director Yan Can submitted:
281
使 使 [60] [61] [62] [63]
He Chengtian, Crown Prince's Director of Standards and Director of the Imperial Academy, petitioned to revise the Yuanjia calendar; lunar-eclipse tests place today's winter solstice at Dipper 17, and earth-standard shadow measurements show winter solstice already three days off. An edict ordered the proposal sent out for verification. In the eleventh year of Yuanjia we were ordered to test lunar eclipses and earth-standard shadows; under the Jingchu method then in use, winter solstice placed the sun slightly past Dipper 21. For the full-moon eclipse on day 16 of month 7, year 11, the appointed hour was mao; the eclipse began on the 15th at the start of the second watch, fourth chou drum, and ended at the fourth drum in the last degree of Encampment 15. Jingchu placed the sun that day at Chariot 3. By opposition, the sun should have stood at Wings 15½. [60] Again, for the full-moon eclipse on day 16 of month 12, year 13, the appointed hour was you; the eclipse began at the start of hai and ended at the first watch, third drum, in Ghost 4. Jingchu placed the sun that day at Woman 3. By opposition, the sun should have stood at Ox 6½. Again, for the full-moon eclipse on day 16 of month 12, year 14, [61] the appointed hour was mid-xu; the eclipse began at the second watch, fourth drum, near the end of hai, and ended at the third watch, first drum, in Well 38. [62] Jingchu placed the sun that day at Dipper 25. By opposition, the sun should have stood at Dipper 22½. [63] For the full-moon eclipse on day 15 of month 5, year 15, the appointed hour was xu; the moon had barely risen when a quarter of the disk was already eclipsed, near Dipper 16. Jingchu placed the sun that day at Well 24. By opposition, the sun should have stood at Well 20. Again, for the full-moon eclipse on day 16 of month 9, year 17, the appointed hour was just past zi; on the 15th, before the second watch, first drum, the eclipse began; at the third drum twelve-fifteenths of the disk was eclipsed in Hairy Head 1½. Jingchu placed the sun that day at Room 2. By opposition, the sun should have stood at Base 13½. Across all five eclipses, opposition at 182½ degrees shows winter solstice not at Jingchu's Dipper 21+ but consistently near Dipper 17½—exactly as He Chengtian claimed.
282
[64] [65]
From year 11 onward we also measured shadows with the earth-standard. That year Jingchu placed winter solstice on month 11, day 7; clouds before and after hid the shadow. Winter solstice of year 12 fell on month 11, day 18; the longest shadow came on day 15. Winter solstice of year 13 fell on month 11, day 29; the longest shadow came on day 26. Winter solstice of year 14 fell on month 11, day 11; clouds before and after hid the shadow. [64] Winter solstice of year 15 fell on month 11, day 21; the longest shadow came on day 18. Winter solstice of year 16 fell on month 11, day 2; the longest shadow came on month 10, day 29. Winter solstice of year 17 fell on month 11, day 13; the longest shadow came on day 10. Winter solstice of year 18 fell on month 11, day 25; the longest shadow came on day 21. [65] Winter solstice of year 19 fell on month 11, day 6; the longest shadow came on day 3. Winter solstice of year 20 fell on month 11, day 16; clouds before and after hid the shadow. Comparing all years, the longest shadow consistently precedes the computed winter solstice by three days. Lunar-eclipse tests already show the sun four degrees off. Earth-standard shadows show winter solstice three days off as well. Today's true winter solstice falls near Dipper 14—again matching He Chengtian.
283
He Chengtian's method also fixes large and small remainders for every new moon, full moon, and quarter; though syzygy times are carefully computed, the surplus-deficit correction produces runs of three long or two short months—quite unlike the old system. Under the old method, solar eclipses occurred not only at new moon but also at month-end and on the second day. This is what the Gongyang Commentary means by "sometimes early, sometimes late." I hold that this one rule should remain unchanged.
284
退便
Pi Yanzong, Staff Master for Miscellaneous Cavalry, also challenged He Chengtian: "If new and full moons fix the remainders and the era head lands in surplus, one day is dropped and the old year's last day should become the new era head." He Chengtian therefore revised the new method back toward the old procedure, no longer fixing monthly remainders, as Yanzong urged and the Grand Astrologer reported.
285
The relevant offices memorialized: "Calendar reform is a great state rite; since Han and Wei it has been revised again and again. No method stays correct forever; one adopts what fits the age. Now that the imperial design shines forth and the realm is restored, the gnomon and celestial degrees should be verified to proclaim renewal. He Chengtian's calendar is fit for adoption. In Song year 22 the Yuanjia calendar was adopted throughout the realm." The edict approved.
286
Collation Notes
287
On "Impeaching Shouwang for opposing Heaven and Earth": all earlier editions read the cited text above the cited text; deleted per the Han Treatise. The Han Treatise reads "Heaven's Way" for "Heaven and Earth."
288
On Emperor Zhang summoning Li Fan and fellow calendar compilers: all editions read the cited text; changed to the cited text per the Continuation of the Han Treatise.
289
On "the sun at twenty-one degrees of Dipper": all editions read the cited text; changed to the cited text per Lu Wenchao in the Collected Commentaries on the Later Han.
290
On Yanguang year 3 of Emperor An: the Continuation of the Han Treatise reads "year 2."
291
On "accumulated over many years in mutual succession": all editions omit the cited text; supplied per the Jin Treatise.
292
On "mid-year": all editions also have the cited text above the cited text. Ancient calendars had no term "era-day mid-year"; under the Jingchu calendar entry in Kaiyuan Occupations Classic 105 it reads simply "mid-year" without "era-day"—deleted accordingly.
293
Communication total 790,110: all editions read the cited text for the cited text; changed per the Bureau edition and Jin Treatise.
294
滿 滿
On "at era head, syzygy moon inside the sun's path": the procedure below adds nodal era difference forward to obtain the next era's nodal rate. When the sum fills the communication total, subtract it; the moon is then outside the sun's path. This era and the following Jiayin era nodal rates are all remainders after removing full communication totals. Every syzygy at this era head should therefore read "moon outside the sun's path."
295
On "if below the nearest median qi the interval limit is less than the limit number": the text requires the cited text below the cited text.
296
滿滿
On carrying from small fraction and small remainder: all editions omit the cited text below the cited text; supplied per the Jin Treatise.
297
Interval limit 1,147: all editions read the cited text; changed to the cited text per the Bureau edition and Jin Treatise.
298
Limit number 1,122: all editions read the cited text; now following the Bureau edition.
299
Interval limit 1,036: all editions read the cited text; the Jin Treatise wrongly reads the cited text; now following the Bureau edition.
300
Limit number 823: all editions read the cited text; changed to the cited text per the Bureau edition and Jin Treatise.
301
Interval limit 812: all editions and the Jin Treatise read the cited text; now following the Bureau edition.
302
Interval limit 801: all editions omit the cited text; supplied per the Bureau edition and Jin Treatise.
303
Interval limit 1,157: all editions omit the cited text; supplied per the Bureau edition and Jin Treatise.
304
Limit number 1,181: all editions read the cited text; supplied as the cited text per the Bureau edition and Jin Treatise.
305
廿
On "in the last ten days of winter, the moon at Heart of Net": all editions read the cited text for the cited text; except the Bureau edition, all read the cited text for the cited text; now changed per the Continuation of the Han Treatise and Qian Daxin.
306
滿
On carrying large fraction from new-moon midnight sun-degree: all editions omit the cited text below the cited text; supplied per the sense of the text.
307
On nodal departure fraction like nodal syzygy total or below: all editions read the cited text for the cited text; changed per the Jin Treatise.
308
On nodal syzygy and lunar eclipse like nodal syzygy total or below: all editions read the cited text for the cited text; changed per the Jin Treatise.
309
On dividing nodal departure degree by the day divisor: the cited text should read the cited text by the sense of the text.
310
On removing nodal departure degree fraction: all editions omit the cited text; supplied per the Jin Treatise.
311
Surplus initial: all editions read the cited text; changed to the cited text per the Bureau edition and Jin Treatise.
312
271: all editions omit the cited text; supplied per the Bureau edition and Jin Treatise.
313
Deficit accumulated parts 278,099: all editions read the cited text for the cited text. Changed per the Bureau edition and Jin Treatise.
314
Day remainder 4,450: all editions omit the cited text; now following the Bureau edition.
315
On the day remainder entered in the calendar: all editions omit the cited text; supplied per the Jin Treatise.
316
On multiplying deficit rate by calendar day remainder: all editions read the cited text for the cited text; transposed per the Bureau edition and Jin Treatise.
317
沿
Bond 6 〈Half-strong.〉 All editions read "Bond 5 〈Half-strong.〉" ; erroneous—now corrected. The Jingchu calendar's twenty-four qi values largely follow the Quarter-Remainder system; minor numerical differences reflect slight dipper-fraction variation between the two. Table figures are computed by the method in Li Rui's commentary on the Quarter-Remainder procedure. Large discrepancies below are corrected in the text. Minor tail-digit differences are noted without changing the received text.
318
Emptiness 5 〈Half-weak.〉 All editions read "Maiden 〈Half-strong.〉" ; erroneous—now corrected.
319
Stomach 11 〈Very strong.〉 Should read "Stomach 11 〈Half-strong.〉"
320
Winnowing Basket 〈Half-weak.〉 Should read "Winnowing Basket 〈Half-strong.〉"
321
Net 6 〈Very.〉 Should read "Net 7."
322
Tail 15 〈Half-strong.〉 Should read "Tail 15 〈Half-weak.〉"
323
Five chi five cun: all editions read "5 chi 5 cun 〈2 fractional parts.〉" ; erroneous—now deleted and corrected.
324
Neck 8 〈Half-weak.〉 Should read "Neck 8 〈Shao-weak.〉"
325
1 zhang: all editions read "1 zhang 8 cun 〈3 fractional parts.〉" ; erroneous—now corrected.
326
Room 3 〈Half-strong.〉 Should read "Room 3 〈Very strong.〉"
327
Wings 15 〈Very.〉 Should read "Wings 15 〈Very weak.〉"
328
45 〈5 fractional parts.〉 The Sanchao edition correctly reads "5 fractional parts." All editions wrongly read "3 fractional parts."
329
Chariot 15 〈Shao-strong.〉 Should read "Chariot 15 〈Shao.〉"
330
On subtracting from each qi's meridian stars: all editions omit the cited text below the cited text; supplied per the Jin Treatise. The character the cited text is probably omitted below the cited text.
331
Year-count year then is called total cycle year-count; year end then is called total cycle conjunction-count: by the sense of the text it should read "year then is called total cycle year-count, end then is called total cycle conjunction-count.
332
On the new-moon small remainder: all editions omit the cited text; supplied per the Jin Treatise.
333
On reducing by the communication divisor: all editions add superfluous the cited text below the cited text; deleted per the Jin Treatise.
334
On day entered in month remainder: all editions omit the cited text below the cited text; now following the Bureau edition.
335
Degree remainder 1,472,869: all editions omit the cited text; supplied per the Bureau edition and Jin Treatise.
336
Day remainder 20,344,261: all editions omit the cited text above the cited text; now supplied.
337
On "auxiliary, multiply by rule intercalations": the cited text is probably superfluous.
338
Small remainder 2,419: all editions read the cited text for the cited text; changed per the Bureau edition and Jin Treatise.
339
On seeking the next day entered in month: the cited text should follow the cited text.
340
On adding day entered in month and remainder: the cited text should follow the cited text.
341
Remove dipper fraction: by the sense it should read "passing the Dipper, remove dipper fraction."
342
On Ji Han's Preface to the Fu on the Road Spirit: all editions read the cited text; changed per Shen Tao. Shen Tao's Random Notes from the Bronze Iron notes: "This is Ji Han's Preface to the Fu on the Road Spirit in Initial Learning Record 13, Ritual Section. "the cited text" corrupts "the cited text", and "the cited text" and "the cited text" were lost in transmission."
343
Cui Shi's Monthly Ordinances for the Four Peoples says 〈To.〉 On "that the multitude of ancestors may come and draw near": Zhang Yuanji says it does not connect with the preceding text and is a displaced passage from the Treatise on Rites. Sun Biao's Study of the Song History argues that this road-spirit section does not belong in the calendar treatise. Also, "Four Peoples" originally read "Four Men"—presumably altered to avoid Tang taboo; now corrected.
344
廿
On Liu Zhi of Pingyuan under Emperor Wu of Jin: all editions read "Eastern Jin time"; changed per the Jin Treatise. Qian Daxin notes: "Liu Zhi, styled Zifang, was younger brother of Minister of Works Shi. He served Emperor Wu, not Eastern Jin—the Treatise errs."
345
On Heaven's southern extreme: all editions omit the cited text; supplied per the Comprehensive Mirror, the twenty-first year of Yuanjia.
346
The sun that day should have stood at fifteen and a half degrees of Wings: all editions read "Wings fifteen and a half degrees." Since the eclipse ended in Encampment 15, opposition at 182° places the sun at Wings 16½, not 15½.
347
()
On the full-moon eclipse of month 12, day 16, year 14: all editions read "twelfth month." From the thirteenth year of Yuanjia month 12 full moon to the fourteenth year of Yuanjia month 12 full moon exceeds one eclipse year; a lunar eclipse is impossible. Recalculation shows a dinghai full moon in month 11 of that year (day 16) lunar eclipse; the received text is wrong.
348
On "at Well 38": all editions read "38 degrees" without naming Well. Well has only 33 degrees; the received figure is clearly wrong. Recalculation places the full-moon eclipse in the eleventh month of the fourteenth year of Yuanjia at Well 26.
349
The sun that day should have stood at twenty-two and a half degrees of Dipper: all editions read "twenty-two and a half degrees." Testing these eclipses against Jingchu shows a 3½° error; Jingchu places the sun at Dipper 25 that day, but the true position is Dipper 21½. Recalculation places that month's full-moon eclipse at Well 26; opposition agrees. The text should therefore read "21½ degrees."
350
On "overcast skies showed none": following parallel passages, the cited text should follow the cited text.
351
On "twenty-first day the shadow was longest": all editions read only "twenty-first day." Following parallel cases, earth-standard shadows place the longest shadow three days before computed winter solstice; with solstice on day 25, the longest shadow should fall on day 22.
← Previous Chapter
Back to Chapters
Next Chapter →