The Five Surviving Editions of the Lishui County Gazetteer

The five surviving editions

A report by Jiangsu News states that only five editions of the Lishui county gazetteer survive: those of the Wanli, Shunzhi, Kangxi, Qianlong and Guangxu periods. The report does not give compilers or extent for each, so the entries below are cross-checked against the individual catalogue cards, the note in the local gazetteer portal and the antiquarian catalogue entry. Where a statement rests on a single catalogue, it is marked as such in the text.

Edition	Extent	Compilers	Edition
Wanli	8 juan plus a prefatory juan	revised by Wu Shiquan, compiled by Huang Rujin	first printed in 1579; the surviving impression is the 1620 supplement
Shunzhi	10 juan with 1 prefatory juan	revised by Min Pailu, compiled by Lin Gudu	woodblock edition of 1658
Kangxi	11 juan with 1 prefatory juan	revised by Liu Dengke; compiled by Cheng Zhiwang, Xie Wenyun, Wang Zhizao and others	woodblock edition of 1677
Qianlong	16 juan plus a prefatory juan	revised by Ling Shiyu; compiled by Fang Xingcun and others	woodblock edition of 1777
Guangxu	22 juan with 1 prefatory juan	revised by Fu Guangguang, compiled by Ding Weicheng	woodblock edition of the Guangxu period; 16 headings and close on a hundred sections

Beyond these five, the two earlier compilations — the first, of 1509, by Chen Xian and Fan Qi, and the second, of 1525, by Wang Congshan, Fang Yan and Huang Zhida — have no surviving copies. They are known only from what the Wanli edition records and from the judgements of later writers.

The Wanli edition: the earliest, and the one with two dates

The Wanli edition is the earliest of the five and the one with the most tangled bibliographical history. Two dates must be kept apart: it was first printed in the seventh year of the Wanli era (1579), while the only surviving impression that can be traced is the supplement of the first year of the Taichang era (1620). According to the local gazetteer portal, that single traceable impression is the copy from the Mao family's Jiguge library, formerly in the Jia collection of the National Peiping Library and now in the National Palace Museum in Taipei; the copies held by the National Library of China, the Shanghai Library and the Nanjing Library are mostly microfilm. The portal adds that "only one impression is known to survive" and "three libraries hold microfilm" are two separate matters and must not be run together.

The occasion of the compilation and its arrangement are given in the antiquarian catalogue entry: in the fourth year of the Wanli era the prefecture of Yingtian recompiled its own gazetteer and called on the subordinate counties to gather up local materials that had been lost, whereupon Wu Shiquan, the magistrate, asked Huang Rujin and other local men to compile this edition; the prefatory juan carries Wu Shiquan's preface of 1579, Wu Shanggeng's preface, and Huang Rujin's short introduction to the new gazetteer. The same catalogue entry gives the eight juan as: juan 1, annals, administrative history, officials, examination success; juan 2, schools, public offices, sacrifices; juan 3, customs, land tax; juan 4, boundaries, city moat, mountains and waters, bridges; juan 5, markets and towns, lanes, antiquities, tombs; juan 6, notable officials, the loyal, local worthies, filial sons; juan 7, righteous men, recluses, virtuous women, sojourners, clergy; juan 8, literature. That catalogue is a private antiquarian site, not an official bibliographic institution, and this site uses it only for the compilers' names and the extent; the headings have not been checked against the original, and any citation of them must say "according to the catalogue".

The Shunzhi edition: the earliest of the Qing

The Shunzhi edition follows the arrangement of the Wanli edition with additions and omissions. It contains an annalistic section, tables of administrative history, officials, education officers and military examination success, an inquiry into recommendation, tables of the examinations, and sections on institutions, sacrifices, relief, customs, land tax, mountains and waters, and geography, together with the various classes of biography. Min Pailu, the reviser (courtesy name Bozong, a native of Xiangfu in Henan, a selected tribute student), was magistrate of Lishui between 1654 and 1656; the annals of the Kangxi edition record that he rebuilt the county hall, repaired the granary at Honglan Wharf and compiled a gazetteer of the city god's temple, and that he drew up the Record of Equalised Corvée. The compiler, Lin Gudu (courtesy name Maozhi, style name Naezi, also known as the Daoshi of Rushan, a native of Fuqing in Fujian), was a poet of the late Ming and early Qing who never sought office and travelled to study at Jinling.

No accessible full text or facsimile of this edition has been found, so this site cites its catalogue information only and not its text. One catalogue gives it wrongly as the 1655 edition and omits the prefatory juan; this is a disagreement between catalogues, and this site sets it out without deciding between them.

The Kangxi edition: the main source for this site's annals

The Kangxi edition runs to eleven juan with a prefatory juan and is catalogued as a woodblock edition of the sixteenth year of the Kangxi era (1677), revised by Liu Dengke and compiled by Cheng Zhiwang, Xie Wenyun, Wang Zhizao and others. Liu Dengke was a native of Fengtian, present-day Shenyang, a provincial graduate, and magistrate of Lishui from 1673; Cheng Zhiwang was the instructor of the county school; Xie Wenyun was a local man and a metropolitan graduate of the Shunzhi period; Wang Zhizao, courtesy name Xingyou, was a local man and a provincial graduate of 1654. One catalogue gives the edition wrongly as 1676 and omits Xie Wenyun.

The first juan of this edition, the annals, is a chronological record running from the Spring and Autumn period to the end of the Shunzhi reign, with each entry carrying both its sexagenary and its corresponding Western year. It is the main source for this site's annals, administrative history and gazetteer selections, and has been read in full. According to the Jiangsu Provincial Local Gazetteers Office, few copies survive in China; there is a colour facsimile of a Japanese copy, and the edition is included in the series of rare local gazetteers held by the National Library of China. Three forms must here be distinguished: the original woodblock impression, the facsimile, and the microfilm. Inclusion in a facsimile series does not indicate where the original impression is held.

The Qianlong edition: the immediate source of the Guangxu edition

The Qianlong edition runs to sixteen juan plus a prefatory juan and is catalogued as a woodblock edition of the forty-second year of the Qianlong era (1777), revised by Ling Shiyu and compiled by Fang Xingcun and others. Ling Shiyu, courtesy name Shuchao, was a native of Qiantang in Zhejiang, a metropolitan graduate of 1771, and magistrate of Lishui from 1773; Fang Xingcun was a native of Tongcheng in Anhui, and Wu Heling and others took part in the compilation. The prefatory juan carries prefaces of 1777 by Zhang Pangui, Ling Shiyu and Jia Wenxing, and the note records a copy held by the Xujiahui Library. The arrangement was recast, and the edition includes pictures and poems of the Eight Views of Zhongshan and 138 pieces of literary writing. Different catalogues describe its nature as a woodblock edition, a printed edition or a manuscript copy; the original has not been examined, and this site records only that the catalogues disagree.

Its place in the sequence is unusual: the general principles of the Guangxu edition state that "the old gazetteers were compiled six times; of the earlier ones no copies remain, and only the Ling edition of 1777 survives, so this edition has been continued from it, taking the Ling text and adding to it" — that is, the Guangxu edition was made by enlarging this one.

The Guangxu edition: the largest and most complete

The Guangxu edition is the largest and most complete of the five, in twenty-two juan with a prefatory juan, with sixteen headings, twenty-two juan and close on a hundred sections. The prefatory juan carries four prefaces: Zhao Youchen's of the eighth month of the seventh year of the Guangxu era (1881), Fu Guangguang's of the ninth month of the same year, Shi Chungao's of the tenth month of the ninth year (1883), and a preface for a reprint by Tan Rixiang of the fifteenth year (1889). The general principles state that "the county of Lishui was founded in the eleventh year of the Kaihuang era of the Sui, and the tables of its administrative history have all been revised and corrected". This edition is the main gazetteer source for this site's articles on administrative history and the source text for material on antiquities, officials and taxation.

"The old gazetteers were compiled six times" and the five that survive

That sentence in the general principles — six compilations, no copies of the earlier ones remaining — appears at first to conflict with the five editions that survive today, and needs some untangling. Taking the count as it stands, the six would be the compilations of 1509, 1525, 1579, 1658, 1677 and 1777, with the Guangxu edition as the seventh; and the statement that no copies of the earlier ones remain would apply to what could be seen in the county at the time of compilation — the Wanli edition survives in a single impression held far away in Taipei, the Shunzhi and Qianlong editions were long known only from catalogue entries, and few copies of the Kangxi edition survive in China though a Japanese copy has been reproduced in facsimile. No source says any of this. The reading above is this site's own attempt to reconcile the dates, and confidence is accordingly set to medium.

Disagreements and points still to be verified

First, the first printing of the Wanli edition in 1579 and the surviving impression of 1620 are two different things, and the two must not be merged in writing.

Second, the disagreement between 1655 and 1658 for the Shunzhi edition and between 1676 and 1677 for the Kangxi edition are differences between catalogues; this site writes only that the catalogues disagree, and does not say that any one of them is wrong.

Third, the extent and compilers of the five editions come mainly from catalogue entries and have not been checked one by one against the original prefatory juan; where that is so, the entries state "according to the catalogue".

Fourth, the full tables of contents of the five editions, the filiation between them, and the date at which the 1509 and 1525 compilations were lost, all await the original or a facsimile.