← Back to 宋史

卷七十六 志第二十九 律曆九

Volume 76 Treatises 29: Measures and Calendar 9

Chapter 76 of 宋史 · History of Song
← Previous Chapter
Chapter 76
Next Chapter →
1
The Huangyou Armillary Sphere
2
退 殿
Yao charged Xihe with fashioning transverse sight-tubes to measure the stars, and had the mechanisms and pivots made of jade so they would not warp with damp or dryness, would move evenly, and would endure firm and unchanging for long use. In later ages, bronze was cast into a spherical instrument to represent the celestial sphere. From Luoxia Hong's compilation of the Taichu Calendar with the armillary sphere down to the Eastern Han Emperor He, the Grand Astrologer's office possessed only an equatorial instrument, and its seasonal observations drifted noticeably. The emperor asked the awaited-edict astrologers Yao Chong and others, and they all replied: "The star chart follows fixed rules, and the sun and moon truly travel the ecliptic; because we lack the proper instrument today, our readings go astray." In the fifteenth year of Yongyuan (103 CE), Jia Kui first installed an ecliptic instrument. In the seventh year of Yanxi (164 CE) under Emperor Huan, Zhang Heng redesigned it, using quarter-degrees as the unit. Later, Lu Ji, Wang Fan, Kong Ting, Hu Lan, Liang Lingzan, and Li Chunfeng each produced instruments of their own. The turmoil and collapse of the Five Dynasties swept away the transmitted methods entirely. Early in the Xiangfu era under Emperor Zhenzong, Han Xianfu built an armillary sphere in which only the pivoting rings gripping the sight-tube could turn, while the ecliptic and equatorial rings remained fixed. Early in the Huangyou era, the court again ordered the solar officials Shu Yijian, Yu Yuan, and Zhou Cong to draw on the designs of Li Chunfeng and Liang Lingzan, recast an ecliptic armillary sphere, and also build clepsydra graduations and a gnomon table; Hanlin Academician Qian Mingyi was directed to work out the methods, and Inner Attendant Mai Yunyan to supervise construction. When they were finished, the armillary sphere was installed on the observation terrace of the Hanlin Astronomical Bureau, the clepsydra in the bell-and-drum tower of Wende Hall, and the gnomon table at the Directorate of Astronomy. The emperor drafted Essentials of the Armillary Sphere in ten scrolls, reviewing what earlier ages had done well or poorly, but afterward kept the work in the palace archives and never published it. The methods of the ecliptic pivoting instrument are set forth in full below.
3
The first layer is called the Six Harmonies Instrument.
4
Yang meridian double ring: circumference 2 zhang 3 chi 2 cun 8 fen, diameter 7 chi 7 cun 6 fen, width 6 cun, thickness 6 fen. Set upright north and south, each face marked with the full circuit of 365 and a fraction degrees, and the north celestial pole 35 and a fraction degrees above the horizon.
5
沿
Yin latitude single ring: circumference, diameter, and width match the yang meridian double ring; outer thickness 2 cun 5 fen, inner thickness 1 cun 9 fen. It bears the ten heavenly stems, twelve earthly branches, and eight trigram directions to orient the instrument to the ground. A channel rim above circulates water to establish true level.
6
Heaven's constant single ring: circumference 2 zhang 4 cun 6 fen, diameter 6 chi 8 cun 2 fen, width and thickness each 1 cun 2 fen. It bears the ten stems, twelve branches, and corner-hour divisions for measuring double-hours and clepsydra marks, fixed to the yang meridian and yin latitude rings like a shell around an egg.
7
The second layer is called the Three Luminaries Instrument.
8
Xuanji double ring: circumference 1 zhang 9 chi 5 cun 6 fen, diameter 6 chi 5 cun 2 fen, width 1 cun 4 fen, thickness 1 cun. Each face is evenly graduated for the full circuit of 365 and a fraction degrees, with two pivots set opposite the celestial poles.
9
宿
Equatorial single ring: circumference 1 zhang 9 chi 6 cun 8 fen, diameter 6 chi 5 cun 6 fen, width 1 cun 1 fen, thickness 6 fen. It bears the interval degrees of the twenty-eight lunar lodges and the full circuit of 365 and a fraction degrees, mounted on the Xuanji ring.
10
退
Ecliptic single ring: circumference 1 zhang 9 chi 2 fen, diameter 6 chi 3 cun 4 fen, width 1 cun 2 fen, thickness 1 cun. It bears the full circuit of 365 and a fraction degrees, evenly divided into the twenty-four solar terms, seventy-two pentads, sixty-four hexagrams, and three hundred sixty yarrow stalks. It inclines twenty-four degrees to the equator and crosses it, precessing backward by more than one degree each year.
11
退
Lunar path single ring: circumference 1 zhang 8 chi 6 cun 3 fen, diameter 6 chi 2 cun 1 fen, width 1 cun 1 fen, thickness 5 fen. It bears the nodal degrees, set inside the ecliptic ring, inclined six degrees to the ecliptic; at the end of each nodal cycle it regresses along the ecliptic by one and a fraction degrees; all these rings turn within the Six Harmonies layer.
12
The third layer is called the Four Pivots Instrument.
13
西
Xuanshu double ring: circumference 1 zhang 8 chi 2 cun 1 fen, diameter 6 chi 7 fen, width 2 cun, thickness 7 fen. Each face is graduated for the full circuit of 365 and a fraction degrees; gripping the sight arm against the pivot axis, it rotates east and west within the Three Luminaries layer to take star positions.
14
使宿
Transverse sight-tube: five chi seven cun long, square outside and round inside, with a central sight hole six fen across, aligned with the solar disc; mounted in the Xuanshu sight arm, it pivots north-south and tilts up and down to sight any star or lodge.
15
宿
Cross-shaped leveling trough: length 9 chi 4 cun 8 fen, head width 1 chi 2 cun 7 fen, body width 9 cun 2 fen, height 7 chi. The trough is one cun wide and eight fen deep; four pillars, each six chi seven cun eight fen long, stand at the ends of the trough to bear the celestial assembly; all are cast in bronze. In this way it gauges the seven luminaries' distances and phases of waxing and waning, revealing how day and night lengthen and shorten. The interval degrees of the twenty-eight lodges measured with it are recorded below; The stars' lodge entries, polar distances, and the fortunes they govern are set out fully in the Treatise on Astronomy.
16
Horn: 12°; Neck: 9°; Root: 16°; Chamber: 5°; Heart: 4°; Tail: 19°; Winnowing-basket: 10°; Dipper: 25°; Ox: 7°; Maid: 11°; Void: 10°; Rooftop: 16°; Encampment: 17°; Wall: 9°; Legs: 16°; Bond: 12°; Stomach: 15°; Hairy Head: 11°; Net: 18°; Turtle Beak: 1°; Three Stars: 10°; Well: 34°; Ghost: 2°; Willow: 14°; Star: 7°; Extended Net: 18°; Wings: 18°; Chariot Axle: 17°.
17
The Huangyou Clepsydra Graduations
18
殿 調使
From the Yellow Emperor's observation of dripping water and devising instruments to set standards, the Three Dynasties appointed officials to the task—the Qiehu clan held that charge. Later makers devised descending clepsydras, floating clepsydras, wheel clepsydras, or balance-and-weight clocks—the designs varied widely. The Song court already possessed graduated clepsydras and water-balance clocks, kept in the east wing of Wende Hall. In the third year of Jingyou (1036) it was recalibrated, yet the water still ran unevenly; at the responsible office's request, one leveling jar, two siphon birds, and twenty-one day-and-night arrows were added. Yet it routinely took sunrise in each season to signal the first mark of mao proper, and by each double-hour's proper mark one mark had already been struck; by the eighth mark the next double-hour was sounding—so the start and end of successive double-hours overlapped by nearly half. Early in Huangyou, an edict directed Shu Yijian, Yu Yuan, and Zhou Cong to rebuild the clock; their method used a leveling weighted jar to steady the flow and eliminate uneven dripping. The day and night were divided into one hundred marks; at winter solstice, forty marks by day and sixty by night; at summer solstice, sixty marks by day and forty by night; at the spring and autumn equinoxes, day and night each received fifty marks. Two and a half marks before sunrise counted as dawn, and two and a half after sunset as dusk; five marks were shifted from night to day—this was called the dusk-and-dawn clepsydra adjustment. All varied with the solar terms. Between winter and summer solstice the day-night span differed by twenty marks in all; each mark's difference had its own arrow, the sequence starting alternately from winter solstice—forty-one arrows in total. Day was divided into morning, mid-morning, noon, afternoon, and evening; night into the five watches jia through wu; dusk and dawn had culminating stars—each arrow bore its own set of figures. Whenever the ecliptic's elevation changed by two degrees forty minutes, the arrows were adjusted to match the calendar. Each double-hour began with one mark and reached its proper mark at four and one-sixth marks; at eight and two-sixths marks it passed into the next double-hour. Below are tabulated the twenty-four solar terms, day and night lengths with sunrise and sunset marks, and dusk and dawn culminating stars for cross-reference.
19
The Huangyou Gnomon Table
20
To read heaven and earth's yin-yang pattern, fix position and direction, and set seasons and intercalation, nothing serves better than the gnomon table. Under the Song, He Chengtian first erected a gnomon to observe solar shadows; within ten years he found that winter solstice by the old Jingchu Calendar regularly lagged the heavens by three days. Later, in Tang times Yixing compiled the Dayan Calendar and, measuring with the gnomon table, found the old calendar's solar terms regularly lagged the heavens by one day. The Directorate's present gnomon table was built by Later Jin astronomical adviser Zhao Yanyi; the table had tilted and the gnomon platform had sunk and shifted, so it could no longer fix celestial degrees accurately. Early in Huangyou, Zhou Cong, Yu Yuan, and Shu Yijian were ordered to rebuild it; studying ancient methods, they erected an eight-chi bronze gnomon two cun thick and four cun wide, joined to a stone platform one zhang three chi long to capture the full winter-solstice shadow, with twin water channels on the face for leveling and graduated inches, cun, and fen on both sides, plus the twenty-four solar terms' shadow lengths measured at Yue Terrace; the instrument was installed at the Directorate of Astronomy. After three years of observation, they found the solar terms lagged the heavens by half a day compared with the old calendar. They then compiled a work in three scrolls entitled New Book of Yue Terrace Gnomon Shadows, reviewing earlier measurements' errors and setting out computational methods in detail. After it was presented to the throne, Hanlin Academician Fan Zhen was directed to write a preface for the record. Zhou Cong held that the shadow dimensions obtained for the twenty-four solar terms agreed more closely with Wang Pu's figures in the Xiande Qintian Calendar than earlier reckonings had. The seasonal variations in shadow length are recorded below for reference and use.
21
Light Snow: Huangyou year 1 (1049), cyclical day wuyin, tenth month, day 19.
22
Measured with the new table: shadow 1 zhang 1 chi 3 cun 5 fen; Wang Pu's calculation: 1 zhang 1 chi 3 cun 9 fen; new method: 1 zhang 1 chi 3 cun 4 fen 48 small fen.
23
Year 2 (1050), day guiwei, tenth month, day 29: overcast—no measurement taken.
24
Year 3 (1051), cyclical day wuzi, tenth month, day 10.
25
Measured with the new table: 1 zhang 1 chi 3 cun; Wang Pu's calculation: 1 zhang 1 chi 4 cun 7 fen; new method: 1 zhang 1 chi 2 cun 9 fen 98 small fen.
26
Heavy Snow: Huangyou year 1, cyclical day guisi, eleventh month, day 4. Overcast—no measurement taken.
27
Year 2, cyclical day wuxu, eleventh month, day 15.
28
Measured with the new table: 1 zhang 2 chi 4 cun 5 fen and a half; Wang Pu's calculation: 1 zhang 2 chi 4 cun 5 fen; new method: 1 zhang 2 chi 4 cun 4 fen 25 small fen.
29
Winter Solstice: Huangyou year 1, cyclical day wushen, eleventh month, day 19.
30
Measured with the new table: 1 zhang 2 chi 8 cun 5 fen; Wang Pu's calculation: 1 zhang 2 chi 8 cun 6 fen; new method: 1 zhang 2 chi 8 cun 5 fen.
31
Year 2, cyclical day guichou, eleventh month, day 30.
32
Measured with the new table: 1 zhang 2 chi 8 cun 4 fen; Wang Pu's calculation: 1 zhang 2 chi 8 cun 6 fen; new method: 1 zhang 2 chi 8 cun 5 fen.
33
Year 3, day jiwei, eleventh month, day 12: overcast—no measurement taken.
34
Slight Cold: Huangyou year 1, cyclical day guihai, twelfth month, day 4.
35
Measured with the new table: 1 zhang 2 chi 4 cun; Wang Pu's calculation: 1 zhang 2 chi 4 cun 8 fen; new method: 1 zhang 2 chi 4 cun 15 small fen.
36
Year 2, intercalary eleventh month, day 15, cyclical day wuchen: overcast—no measurement taken.
37
Year 3, cyclical day jiaxu, eleventh month, day 27.
38
Measured with the new table: 1 zhang 2 chi 3 cun 7 fen; Wang Pu's calculation: 1 zhang 2 chi 4 cun 8 fen 26 small fen.
39
Great Cold: Huangyou year 1, day wuyin, twelfth month, day 19: overcast—no measurement taken.
40
Year 2, cyclical day jiashen, twelfth month, day 1.
41
Measured with the new table: 1 zhang 1 chi 1 cun 7 fen; Wang Pu's calculation: 1 zhang 1 chi 4 cun 4 fen; new method: 1 zhang 1 chi 1 cun 8 fen 40 small fen.
42
Year 3, day jichou, twelfth month, day 12: overcast—no measurement taken.
43
Beginning of Spring: Year 2, day jiawu, first month, day 6: overcast—no measurement taken.
44
Year 3, day jihai, twelfth month, day 16: overcast—no measurement taken.
45
Year 4 (1052), cyclical day jiachen, twelfth month, day 27.
46
Measured with the new table: 9 chi 6 cun 7 fen and a half; Wang Pu's calculation: 1 zhang 1 chi 5 fen; new method: 1 zhang 6 cun 8 fen 7 small fen.
47
Rain Water: Year 2, day jiyou, first month, day 21: overcast—no measurement taken.
48
Year 3, cyclical day jiayin, first month, day 2.
49
Measured with the new table: 8 chi 1 cun and a half fen; Wang Pu's calculation: 8 chi 5 cun; new method: 8 chi 9 cun 76 small fen.
50
Year 4 (1052), cyclical day jiwei, first month, day 12.
51
Measured with the new table: 8 chi 1 cun 2 fen and a half; Wang Pu's calculation: 8 chi 6 cun 1 fen; new method: 8 chi 1 cun 2 fen 18 small fen.
52
Awakening of Insects: Year 2, cyclical day jiazi, second month, day 7.
53
Measured with the new table: 6 chi 6 cun 3 fen; Wang Pu's calculation: 6 chi 8 cun 5 fen; new method: 6 chi 6 cun 3 fen 39 small fen.
54
Year 3, cyclical day jisi, first month, day 17.
55
Measured with the new table: 6 chi 6 cun 5 fen; Wang Pu's calculation: 6 chi 8 cun 5 fen; new method: 6 chi 6 cun 5 fen 68 small fen.
56
Year 4 (1052), cyclical day yihai, first month, day 28: overcast—no measurement taken.
57
Spring Equinox: Year 2, cyclical day jimao, second month, day 23.
58
Measured with the new table: 5 chi 3 cun 5 fen; Wang Pu's calculation: 5 chi 2 cun 7 fen; new method: 5 chi 3 cun 4 fen 77 small fen.
59
Year 3, cyclical day yiyou, second month, day 4: overcast—no measurement taken.
60
Year 4 (1052), cyclical day gengyin, second month, day 14.
61
Measured with the new table: 5 chi 3 cun 1 fen; Wang Pu's calculation: 5 chi 2 cun 7 fen; new method: 5 chi 3 cun 72 small fen.
62
Clear and Bright: Year 2, cyclical day yiwei, third month, day 8.
63
Measured with the new table: 4 chi 2 cun; Wang Pu's calculation: 3 chi 8 cun 9 fen; new method: 4 chi 1 cun 8 fen 61 small fen.
64
Year 3, cyclical day gengzi, second month, day 19: overcast—no measurement taken.
65
Year 4 (1052), cyclical day yisi, second month, day 29.
66
Measured with the new table: 4 chi 2 cun 2 fen; Wang Pu's calculation: 3 chi 9 cun 6 fen; new method: 4 chi 2 cun 1 fen 85 small fen.
67
Grain Rain: Year 2, day gengxu, third month, day 23: overcast—no measurement taken.
68
Year 3, cyclical day yimao, third month, day 4.
69
Measured with the new table: 3 chi 3 cun; Wang Pu's calculation: 2 chi 9 cun 6 fen; new method: 3 chi 2 cun 9 fen 86 small fen.
70
Year 4 (1052), cyclical day gengshen, third month, day 15.
71
Measured with the new table: 3 chi 3 cun 1 fen and a half; Wang Pu's calculation: 3 chi 1 cun; new method: 3 chi 3 cun 1 fen 16 small fen.
72
Beginning of Summer: Year 2, cyclical day yichou, fourth month, day 9.
73
Measured with the new table: 2 chi 5 cun 7 fen; Wang Pu's calculation: 2 chi 3 cun; new method: 2 chi 5 cun 6 fen 28 small fen.
74
Year 3, cyclical day gengwu, third month, day 19.
75
Measured with the new table: 2 chi 5 cun 7 fen and a half; Wang Pu's calculation: 2 chi 3 cun; new method: 2 chi 5 cun 7 fen 42 small fen.
76
Year 4 (1052), cyclical day yihai, third month, day 30.
77
Measured with the new table: 2 chi 5 cun 8 fen and a half; Wang Pu's calculation: 2 chi 3 cun 4 fen; new method: 2 chi 5 cun 8 fen 44 small fen.
78
滿
Lesser Fullness: Year 2, cyclical day gengchen, fourth month, day 24.
79
Measured with the new table: 2 chi 3 fen; Wang Pu's calculation: 1 chi 8 cun 6 fen; new method: 2 chi 3 fen 51 small fen.
80
Year 3, cyclical day yiyou, fourth month, day 5.
81
Measured with the new table: 2 chi 3 fen and a half; Wang Pu's calculation: 1 chi 8 cun 6 fen; new method: 2 chi 3 fen 51 small fen.
82
Year 4 (1052), cyclical day xinmao, fourth month, day 16: overcast—no measurement taken.
83
Grain in Ear: Year 2, cyclical day yiwei, fifth month, day 9.
84
Measured with the new table: 1 chi 6 cun 9 fen; Wang Pu's calculation: 1 chi 6 cun; new method: 1 chi 6 cun and a half fen, 97 small fen.
85
Year 3, cyclical day xinchou, fourth month, day 21.
86
Measured with the new table: 1 chi 6 cun 7 fen; Wang Pu's calculation: 1 chi 5 cun 9 fen; new method: 1 chi 6 cun 7 fen 84 small fen.
87
Year 4 (1052), cyclical day bingwu, fifth month, day 2.
88
Measured with the new table: 1 chi 6 cun 8 fen and a half; Wang Pu's calculation: 1 chi 6 cun; new method: 1 chi 6 cun 8 fen 20 small fen.
89
Summer Solstice: Year 2, cyclical day xinhai, fifth month, day 25.
90
Measured with the new table: 1 chi 5 cun 7 fen and a half; Wang Pu's calculation: 1 chi 5 cun 1 fen; new method: 1 chi 5 cun 7 fen.
91
Year 3, day bingchen, fifth month, day 7: overcast—no measurement taken.
92
Year 4 (1052), cyclical day xinyou, fifth month, day 17.
93
Measured with the new table: 1 chi 5 cun 7 fen; Wang Pu's calculation: 1 chi 5 cun 1 fen; new method: 1 chi 5 cun 7 fen.
94
Lesser Heat: Year 2, day bingyin, sixth month, day 11: overcast—no measurement taken.
95
Year 3, cyclical day xinwei, fifth month, day 22.
96
Measured with the new table: 1 chi 6 cun 9 fen and a half; Wang Pu's calculation: 1 chi 6 cun; new method: 1 chi 6 cun 9 fen 75 small fen.
97
Year 4 (1052), day bingzi, sixth month, day 3: overcast—no measurement taken.
98
Greater Heat: Year 2, cyclical day xinsi, sixth month, day 26.
99
Measured with the new table: 2 chi 4 cun; Wang Pu's calculation: 1 chi 8 cun 5 fen; new method: 2 chi 4 fen 97 small fen.
100
Year 3, cyclical day bingxu, sixth month, day 7.
101
Measured with the new table: 2 chi 2 fen excess; Wang Pu's calculation: 1 chi 8 cun 5 fen; new method: 2 chi 4 fen 24 small fen.
102
Year 4 (1052), cyclical day renchen, sixth month, day 19.
103
Measured with the new table: 2 chi 5 fen; Wang Pu's calculation: 1 chi 8 cun 7 fen; new method: 2 chi 6 fen 53 small fen.
104
Beginning of Autumn: Year 2, cyclical day bingshen, seventh month, day 11.
105
Measured with the new table: 2 chi 5 cun 9 fen; Wang Pu's calculation: 2 chi 2 cun 9 fen; new method: 2 chi 5 cun 9 fen 51 small fen.
106
Year 3, cyclical day renyin, sixth month, day 23.
107
Measured with the new table: 2 chi 6 cun 1 fen and a half; Wang Pu's calculation: 2 chi 3 cun 3 fen; new method: 2 chi 6 cun 2 fen 73 small fen.
108
End of Heat: Year 2, day renzi, seventh month, day 27: overcast—no measurement taken.
109
Year 3, cyclical day dingsi, seventh month, day 9.
110
Measured with the new table: 3 chi 3 cun 6 fen; Wang Pu's calculation: 3 chi; new method: 3 chi 3 cun 6 fen 65 small fen.
111
Year 4 (1052), day renxu, seventh month, day 19: overcast—no measurement taken.
112
White Dew: Year 2, day dingmao, eighth month, day 13: overcast—no measurement taken.
113
Year 3, day renshen, seventh month, day 24: overcast—no measurement taken.
114
Year 4 (1052), day dingchou, eighth month, day 5: overcast—no measurement taken.
115
Autumn Equinox: Year 2, day renwu, eighth month, day 28: overcast—no measurement taken.
116
Year 3, cyclical day dinghai, eighth month, day 9.
117
Measured with the new table: 5 chi 3 cun 8 fen; Wang Pu's calculation: 5 chi 2 cun 1 fen; new method: 5 chi 3 cun 8 fen 69 small fen.
118
Year 4 (1052), day renchen, eighth month, day 20: overcast—no measurement taken.
119
Cold Dew: Year 2, day dingyou, ninth month, day 13: overcast—no measurement taken.
120
Year 3, cyclical day renyin, ninth month, day 24.
121
Measured with the new table: 6 chi 6 cun 7 fen; Wang Pu's calculation: 6 chi 8 fen; new method: 6 chi 6 cun 7 fen 88 small fen.
122
Year 4 (1052), cyclical day wushen, ninth month, day 6.
123
Measured with the new table: 6 chi 7 cun 3 fen and a half; Wang Pu's calculation: 6 chi 9 cun 1 fen; new method: 6 chi 7 cun 4 fen 84 small fen.
124
Frost Descent: Year 2, cyclical day renzi, ninth month, day 28.
125
Measured with the new table: 8 chi 1 cun 6 fen; Wang Pu's calculation: 8 chi 4 cun 5 fen; new method: 8 chi 1 cun 4 fen 70 small fen.
126
Year 3, day wuwu, ninth month, day 10: overcast—no measurement taken.
127
Year 4 (1052), cyclical day guihai, ninth month, day 21.
128
Measured with the new table: 8 chi 2 cun; Wang Pu's calculation: 8 chi 5 cun 6 fen; new method: 8 chi 1 cun 9 fen 66 small fen.
129
Beginning of Winter: Year 2, cyclical day wuchen, tenth month, day 14.
130
Measured with the new table: 9 chi 8 cun and a half fen; Wang Pu's calculation: 1 zhang 1 chi; new method: 9 chi 8 cun 1 fen 25 small fen.
131
Year 3, cyclical day guiyou, ninth month, day 25.
132
Measured with the new table: 9 chi 7 cun 9 fen; Wang Pu's calculation: 1 zhang 1 chi; new method: 9 chi 7 cun 8 fen 63 small fen.
133
Year 4 (1052), cyclical day wuyin, tenth month, day 6.
134
Measured with the new table: 9 chi 7 cun 6 fen; Wang Pu's calculation: 1 zhang 1 chi; new method: 9 chi 7 cun 6 fen 10 small fen.
135
Correcting the Time of Shadow Measurement (Early and Late)
136
退調退 退 滿
In the third year of the Xiaping era of Later Han, the Monograph on the Quarter-remainder Calendar records the mid-term shadow at Start of Winter as 1 zhang and at Beginning of Spring as 9 chi 6 cun. At the winter solstice the sun stands at its southern limit and the gnomon shadow is longest. The two qi lie equally many days from the solstice, so their mid-term shadows should match; instead the earlier is longer and the later shorter—a sudden discrepancy of 4 cun. This is proof from the calendar's shadow records that the winter solstice was placed late relative to Heaven. The two mid-term shadows differ by a daily rate of nine and a half fen minus; with even advance and retreat there is little surplus or shortfall. By that rate, if each qi is moved back two days and twelve ke, the shadow lengths would shorten at Start of Winter and lengthen at Beginning of Spring by 2 cun each—both mid-term shadows becoming 9 chi 8 cun, which would be the true days of Start of Winter and Beginning of Spring. By the same reasoning, the calendar's winter solstice also lagged Heaven by two days and twelve ke. In Xiaping year 3, the calendar then in use fixed the winter solstice on dingchou, with the added hour exactly at midday. Subtract two days and twelve ke, and the winter solstice falls on yihai, with the added hour twenty-eight ke after midnight. The Monograph of Song records that in the fifth year of Daming, tenth month, day 10, the shadow was 1 zhang 7 cun 7 fen and a half; on the twenty-fifth day of the eleventh month, 1 zhang 8 cun 1 fen excess; and on the twenty-sixth, 1 zhang 7 cun 5 fen strong. Averaging these, the true midwinter solstice should fall on the third day of the eleventh month—whence one may determine how early or late it was. Subtract the shadow of the two preceding days from that of the two following days; the difference is the daily rate, which doubled gives the divisor. Subtract the shadow of the two earlier days, multiply by one hundred ke, and the product is the dividend. Dividing the dividend by the divisor yields winter solstice's added hour at thirty-one ke after midnight—one day later than the Yuankai Calendar, which matches the true count of celestial days. Checked over many years, the additions and subtractions balance out evenly. Compared across different years, near and far dates align with the expected rate. Reviewing the two schools' theories shows them to be sketchy and incomplete. The Xiping method aimed at a mean but missed the surplus before and after the solstices. The Daming method balanced the left and right rates but erred in setting the dividend and divisor. Comparing shadows and fixing the solar terms are the calendrist's most pressing duties. Earlier comparisons relied only on a few days before and after winter solstice to fix whether the added hour came early or late. Moreover, the shadow's rate of change around the two solstices advances and retreats within the finest gradations. The sun's motion also varies, waxing and waning slightly; if that variation is taken as the standard, the added hour will be misaligned. Jin and Han calendar methods often averaged gnomon shadows measured before and after a term—an error of more than half a day. Year-by-year comparison shows that at Beginning of Winter and Beginning of Spring the shadow shifts by more than an inch; to fix the added hour, the nearest observations should be combined and halved to obtain the general day from solstice; subtract the two gnomon readings; multiply the remainder by the divisor, and when it fills the day's gnomon difference, the quotient is one mark; use the mark-difference to find winter solstice: if the preceding gnomon reading is greater, subtract; if less, add; reverse the procedure for summer solstice.
137
Add or subtract the general day from solstice to obtain the fixed day; add half a day's marks, assign from the preceding interval day's double-hour, and count outward to obtain the two solstices' added hour, double-hour, and marks. Pursued in this way, whether the two solstices' added hour comes early or late can be verified.
138
The Huangyou Yue Terrace Gnomon Shadow Method
139
使 退退 滿滿退 退
The Dayan Calendar's sun carriage and the Chongtian Calendar's fixed-difference rate, though called thoroughly precise, could not fully capture the principle of mutual correspondence above and below, so gnomon degrees could not be made to agree. A new method is now established to match the sun's waxing and waning above and the gougu ratios below, so calculated dimensions and celestial measurements agree without discrepancy. The method runs: count the days after the two solstices, subtract the solstice approximate remainder, and add half a day's fraction to obtain the sought day's noon accumulated number; set this to find the advance-retreat difference—if the accumulated number is ninety-one days thirty-two fen or less (one quadrant), it falls in the front section; If it exceeds one quadrant, subtract it from the two-solstice limit of one hundred eighty-two days sixty-one fen; the remainder falls in the rear section. Place the front and rear degrees above and two hundred below; subtract the lower from the upper, multiply the remainder by both values, divide by four thousand one hundred thirty-five to obtain fen, and any remainder becomes small fen. After winter solstice the result is an advance difference; after summer solstice, a retreat difference.
140
Next set the initial and final limits: if the sought day's noon accumulated number, counted from winter solstice in the initial limit or summer solstice in the final limit, is forty-five days sixty-two fen or less, it falls within the initial limit; If it exceeds that, subtract from the two-solstice limit; the remainder enters the final limit. For the period after winter solstice's final limit and after summer solstice's initial limit, use one hundred thirty-seven days as the rate.
141
退 滿滿 退 滿滿滿 滿
This is used to find the noon gnomon reading. To find the noon gnomon reading, consider which limit applies. If within winter solstice's initial limit or summer solstice's final limit, subtract the limit-entry days from one thousand nine hundred thirty-seven and a half; the remainder is the general difference; multiply the limit days and fractional parts by the advance-retreat difference, reduce by the five-times-one-hundred factor, and subtract from the general difference to obtain the fixed difference; square the limit-entry days and parts, multiply by the fixed difference, and convert at one million per chi (with remainder as cun, fen, and small fen); subtract from the winter solstice constant gnomon of one zhang two chi eight cun five fen to obtain that day's noon gnomon reading. If the date falls in winter solstice's final limit or summer solstice's initial limit, reduce the limit-entry days and parts by one-third, subtract from four hundred eighty-five and a fraction, and the remainder is the general difference; subtract the advance-retreat difference from the extreme number; if the date falls after spring equinox and before autumn equinox, quarter the remainder and add it to the general difference to obtain the fixed difference; If before spring equinox or after autumn equinox, multiply the days and parts from the equinox, divide by six hundred, subtract from the general difference to get the fixed difference; square the limit-entry days and parts, multiply by the fixed difference, convert at one million per chi, and add to the summer solstice constant gnomon of one chi five cun seven fen to obtain that day's noon gnomon reading. With the annual-cycle calendar, multiply that day's gnomon wax-wane difference by the noon remainder, reduce by the divisor, and adjust the base gnomon reading to obtain the day's fixed noon gnomon.
142
Pursued in this way, the principle of mutual correspondence above and below and the basis of gougu oblique projection can all be checked against observation; the full annual-cycle calculation tables for Yue Terrace gnomon shadows follow.
← Previous Chapter
Back to Chapters
Next Chapter →