繁體中文
嘉慶《新安縣志》卷之二 輿地畧一「都里」喺官富司客籍段之後，緊接另一司嘅名單：掃描本第 65 頁下葉見【福永司管屬村莊】標目，標目之後 4 個村名（水貝村、合水口、馬山頭、田寮下）到該葉末為止；本籍段跟住喺第 66 頁上下兩葉、第 67 頁上下兩葉（每葉各 36 名＝9 欄、每欄上下兩個村名、左右兩半）同第 68 頁上葉標目之前 6 名（倉邊村、福永村、下莆村、東坑村、柳屋、龍溪村）續完，逐葉點算 4＋36×4＋6＝154 名，同志錄所記福永司本籍 154 條完全吻合。第 68 頁上葉右半由右數第 4 欄就係【福永司管屬客籍村莊 附】標目（「附」以小字附於標目之下；掃描本標目末字同部分村名用通俗寫法，本站正文一律寫成繁體「村莊」、「黃家莊」），客籍名單 31 名：標目左邊至版心 10 名（官田、蕉莆村、塘頭圍、逕貝村、渭江村、坭岡、白坑、屋場排、羅租村、白芒村）、版心左邊 18 名（黃金洞、黃家莊、黃田村、橫村圍、塘坑圍、石隆、水坑圍、丫坑圍、亞婆髻、龍門村、黃蔴莆、蔗竹塘、蔗園莆、長坑圍、草尾萌、青山下、應人石、藤弔圍），再加第 68 頁下葉 3 名（龍岡仔、龍塘圍、案山村），合共 10＋18＋3＝31 名，同志錄所記福永司客籍 31 條亦完全吻合——本籍 154、客籍 31 兩個數字同時對得上，係「每個掃描頁＝上下兩葉、每葉右半先讀、同一欄由上而下兩個村名」呢套讀法嘅獨立驗證。地理上，可辨認嘅村名全部落在今日深圳：寶安（沙井村、衙邊村、固戍村〔掃描本作「固戌」〕、懷德村、黃田村、新橋村、上寮村、長圳村、水尾圍、塘尾圍、洪田村、茅洲山、塘下涌）、光明（樓村、田尾、西田村、薯田村、塘下村）、石岩同龍華（官田、羅租村、赤坎村、上輋村、下輋村、香園村）、南山（白芒村）——換言之，福永司段並唔係香港新界西（屯門、龍鼓灘、掃管笏、大欖涌）嘅村名單；今日香港境內嘅村落要去官富司本籍段（或縣丞、典史兩段）搵，係下一批嘅掃描目標。同名村落兩地皆有嘅（隔田村、田心村、白沙村、香園村、塘尾圍、田寮村等）一律唔逐村指派司屬；客籍段嘅「青山下」（疑屯門青山腳下一帶，亦可能係深圳同名村落）同「白芒村」（大嶼山白芒有考古遺址同志錄條目，但同段其餘村名盡屬深圳，亦可能係深圳南山白芒）兩個都屬「歧義即棄」，未經覆核前唔會計入任何香港村落數字。段末另有一段跋語（第 68 頁下葉，緊接「案山村」之後）：先講舊志分邑為鄉三、為都七、為圖五十有七、為村將及五百，而各都多有舛錯、前後無殊者不過十之六七；再記七都長山頭、石潭莆、林村、蓮湖一帶數十餘村仍入東莞，柘磡、角子頭、鑑巷圍、孖凹村等十餘處久廢無存、新舊異名亦難詳辨；最後總說「現查本籍村莊五百七十有奇、客籍村莊二百七十有奇，分隸於縣丞、典史兩巡檢屬下」，方志所以一一開載、客籍亦附於後。字形未高倍確認嘅一批（沙莆村、豪涌村〔疑「蠔涌」〕、蔭下村、大囧村、烏木〔末字未能確認〕、公爵藪、白沙苒、苫田村、壆頭岡、鄧家萌、李松萌、山尾新村〔「山」字特小〕）已列入 /coverage 已知缺口，未完成覆核前唔會當作已核實村名引用。
English
The Jiaqing Xin’an County Gazetteer, juan 2, opens another office’s list straight after the Guanfu Office’s Hakka villages. On the lower leaf of page 65 of the scan stands the heading “Villages under the Fuyong Office”, followed by the last four names on that leaf (水貝村, 合水口, 馬山頭 and 田寮下); the indigenous list then runs on through both leaves of page 66 and both leaves of page 67 (thirty-six names to a leaf — nine columns of two names each, left and right halves) and the six names standing before the next heading on the upper leaf of page 68 (倉邊村, 福永村, 下莆村, 東坑村, 柳屋 and 龍溪村), giving by leaf count 4 + 36 × 4 + 6 = 154 names, exactly the 154 indigenous villages the archive records under the Fuyong Office. The fourth column from the right in the right half of the upper leaf of page 68 carries the heading “Hakka villages under the Fuyong Office, appended” (the word “appended” set small below it; as with the Guanfu heading the scan prints the last character of headings and of some village names in the popular form, which this site always writes in the traditional form, 村莊 and 黃家莊). Its thirty-one names are: ten between the heading and the running head (官田, 蕉莆村, 塘頭圍, 逕貝村, 渭江村, 坭岡, 白坑, 屋場排, 羅租村 and 白芒村), eighteen left of the running head (黃金洞, 黃家莊, 黃田村, 橫村圍, 塘坑圍, 石隆, 水坑圍, 丫坑圍, 亞婆髻, 龍門村, 黃蔴莆, 蔗竹塘, 蔗園莆, 長坑圍, 草尾萌, 青山下, 應人石 and 藤弔圍), and three on the lower leaf of page 68 (龍岡仔, 龍塘圍 and 案山村) — 10 + 18 + 3 = 31, exactly the 31 Hakka villages the archive records under the same office. Both figures tallying at once independently verifies the reading order used throughout this work: one scan page holds two leaves stacked, each leaf is read right half first, and each printed column is read downwards through its two village names. Geographically every identifiable name lies in Shenzhen today: Bao’an (沙井村, 衙邊村, 固戍村 — printed 固戌 — 懷德村, 黃田村, 新橋村, 上寮村, 長圳村, 水尾圍, 塘尾圍, 洪田村, 茅洲山 and 塘下涌), Guangming (樓村, 田尾, 西田村, 薯田村 and 塘下村), Shiyan and Longhua (官田, 羅租村, 赤坎村, 上輋村, 下輋村 and 香園村) and Nanshan (白芒村). The Fuyong Office’s list is therefore not a name-list of western Hong Kong (Tuen Mun, Lung Kwu Tan, So Kwun Wat, Tai Lam Chung); villages inside Hong Kong today must be sought in the indigenous Guanfu list or in the county deputy’s and registrar’s stretches, which are the next pages to read. Names that exist on both sides of the border (隔田村, 田心村, 白沙村, 香園村, 塘尾圍, 田寮村 and others) are never pinned to a place and are assigned to no office; two names in the Hakka list — 青山下 (perhaps the foot of Castle Peak at Tuen Mun, perhaps a same-named Shenzhen village) and 白芒村 (Pak Mong on Lantau has archaeological sites and archive entries, but every other name in the same stretch is a Shenzhen village, and Nanshan in Shenzhen has a 白芒 too) — are dropped as ambiguous and counted in no Hong Kong total until they are re-checked. The section closes on the lower leaf of page 68 with a note printed after 案山村: the old gazetteer divided the county into three townships, seven districts, fifty-seven charts and nearly five hundred villages, yet the districts are much confused and only some six or seven in ten hang together; the several dozen villages of the seventh district along 長山頭, 石潭莆, 林村 and 蓮湖 still belonged to Dongguan, while a dozen or so places such as 柘磡, 角子頭, 鑑巷圍 and 孖凹村 had long fallen out of existence, so that old and new names can hardly be told apart; and it ends by stating that the county then had over 570 indigenous and over 270 Hakka villages, under the county deputy, the registrar and two police offices — hence the setting down of every one in turn, with the Hakka villages appended. A set of names whose glyphs could not be confirmed at high magnification (沙莆村, 豪涌村 — perhaps 蠔涌, 蔭下村, 大囧村, 烏木〔last character unconfirmed〕, 公爵藪, 白沙苒, 苫田村, 壆頭岡, 鄧家萌, 李松萌 and 山尾新村〔where 山 is printed very small〕) is recorded under /coverage as a known gap and is never cited as a confirmed village name until it is re-checked.
出處：掃描本：《廣州大典》第三十五輯·史部方志類·第六十三冊，嘉慶《新安縣志》卷之二 輿地畧一「都里」——「福永司管屬村莊」同「福永司管屬客籍村莊 附」兩段，掃描本第 65 頁下葉（「水貝村」起）至第 68 頁下葉（「案山村」同段末跋語止），共六葉多；版心逐葉實測題「新安縣志 卷之二 都里」，頁次係行草字形、未能高倍確認，按次序推定十三至十八（承接第 63 至 65 頁六葉嘅七至十二）；Commons sha1 ad357f97…。掃描本冇可用文字層，內容一律人手核讀（裁圖工具（唔用 OCR） 全葉裁圖＋左右半葉放大圖）：見 scripts/ocr-pd-gazetteer-scans.swift。Provenance: Scan: Guangzhou Dadian, series 35 (local gazetteers), volume 63 — the Jiaqing Xin’an County Gazetteer, juan 2, “Radiations of the Territory” part 1, “Villages”: the sections headed “Villages under the Fuyong Office” and “Hakka villages under the Fuyong Office, appended”, from the lower leaf of page 65 of the scan (starting at 水貝村) to the lower leaf of page 68 (ending at 案山村 and the closing note that follows it), six leaves and more; each leaf’s running head was read directly as “Xin’an County Gazetteer, juan 2, Villages”, its folio numeral being cursive and not confirmable at high magnification, so folios 13 to 18 are given by sequence, following folios 7 to 12 on the six leaves of pages 63 to 65; Commons sha1 ad357f97…. The scan has no usable text layer, so its content is read by hand (full-leaf crops plus magnified left- and right-half details, 裁圖工具（唔用 OCR）): see scripts/ocr-pd-gazetteer-scans.swift.