2026-10-06

Counting the Water The New Gauge Basin drawing of the Pokfoolum Conduit, Hong Kong, 1871–1877: an artefact-led reconstruction of the headworks of a covered brick conduit, and of the road built on top of it

 Research paper

Counting the Water

The New Gauge Basin drawing of the Pokfoolum Conduit, Hong Kong, 1871–1877: an artefact-led reconstruction of the headworks of a covered brick conduit, and of the road built on top of it

Histsyn· Independent research
Compiled 6 October 2026 · Endnotes and bibliography after The Chicago Manual of Style, 17th ed.

Abstract

This paper identifies and reads a single undated office drawing in the Hong Kong Public Records Office, Hong-kong Waterworks – Pokfoolum Conduit. New Gauge Basin (HKRS1951-1-3, original ref W00239), and places it inside the documentary record of the works it proposes. The sheet names itself, and its black and red labels separate an existing Old Gauge Basin and Old Dam from a Gauge Basin proposed in red at the toe of the reservoir embankment, with outlet pipes and a new six-foot path. Four contemporary descriptions of the same structure survive in four different series — the 1875 Gazette report of the Surveyor-General, a 1875 despatch to the Colonial Office, the 1896 Sessional Paper retrospective, and a China Mail account of July 1878 — and a drawing level of 500.99 feet agrees to within a foot with their “500 foot contour.” What follows reconstructs the works from Price’s 1875 estimate of £9,278. 12. 0 through the tender of May 1876 and completion in 1877 at $62,090.76 under W. Danby, and then follows the conduit through its service life: as a measuring instrument (a basin where the day’s supply was “regulated and registered”), as a line under the city (the 1884 Robinson Road resumption, Chadwick’s 1889–92 zoning), and finally as a road — Conduit Road, 干德道. The paper is explicit about what the sheet does not establish: its date, its authorship, and any claim to be first, earliest or longest.

Keywords: Pok Fu Lam / Pokfoolum Conduit; gauge basin; Hong Kong water supply; J. M. Price; W. Danby; Conduit Road 干德道; HKRS1951-1-3; Public Works Department drawings; colonial infrastructure

1. Introduction: a drawing without a date

A sheet of paper in the Hong Kong Public Records Office carries the title HONG-KONG WATERWORKS. POKFOOLUM CONDUIT. NEW GAUGE BASIN. It is a working drawing of the public works office: ink and wash, twenty feet to the inch, a pasted label reading W 239 in the lower right, a red docket in the upper right. It carries no printed date, no signature and no author’s name, and the archive supplies only a file range of 1871–1877 — a range which brackets the works rather than dating the sheet.1 This paper is an attempt to say what the sheet is, when it belongs, and what it therefore lets us see, using the sheet itself as the first witness and the corpus as the second.2

The project is deliberately narrow. One artefact is read at high resolution before anything is searched, so that no secondary source can tell the reader what the primary object says; the corpus is then inventoried and queried for the structure the sheet draws, not for trivia about it; every number is required to stand on two independent contemporary witnesses or be marked as single-witness; every quotation is pulled from the whole document rather than from a search-result snippet; and every document id cited below was re-fetched through its public viewer link and re-checked against the printed page before it was printed. A final adversarial pass tried to falsify the headline claims, and the results — including a zero-hit exact-phrase search for the sheet’s own title — are recorded in the evidence register in § 10 and Appendix A.3

The argument runs as follows. Section 2 sets out the identification in tiers: what the artefact says about itself, what four independent documents say about the same structure, and what neither establishes. Sections 3 and 4 reconstruct the scheme from the Surveyor-General’s report of 10 July 1875 to the tender of 13 May 1876 and the completion recorded in 1877. Section 5 describes what was actually built, and why the sheet is a detail of its headworks rather than a map of it. Sections 6 and 7 treat the works as a hydrological instrument — the conduit captured wet-season water rather than creating dry-season water, and the gauge basin existed in order to measure. Sections 8 and 9 follow the alignment into the city and then into the street plan: the conduit under Robinson Road, the zone boundary in Chadwick’s scheme, and finally the road named after it. Section 10 records what remains unproven, and section 11 concludes.

2. The sheet and how it identifies itself

The identification is offered in tiers, of which the first three are read off the artefact and the last three lie outside it. Tier 1 is the title block itself: two lines of printed lettering reading HONG-KONG WATERWORKS. and POKFOOLUM CONDUIT. NEW GAUGE BASIN., a scale bar reading SCALE 20 FEET = 1 INCH., and the pasted catalogue label W 239, which matches the archive’s original reference W00239.4 No interpretation is involved in any of this; the words are on the paper.



Figure 1. HONG-KONG WATERWORKS. POKFOOLUM CONDUIT. NEW GAUGE BASIN., ink and wash on paper, 20 feet = 1 inch, sheet mark “W 239”, no date and no signature. Public Records Office (Hong Kong), HKRS1951-1-3, original ref W00239, series HKRS1951 Drawings Relating to Water Works, file dates 1871–1877; digital file HKRS1951-1-3_0001.jpg, used here by research reference.5 Red denotes proposed works: the new gauge basin, its outlet pipes and a new six-foot path. Black and blue denote what already existed, including the Old Gauge Basin and Old Dam.

Tier 2 is the vocabulary of the labels, which separates old from new in words and not merely in colour. In black the sheet records Old Gauge Basin, Old Dam., EMBANKMENT, Valve Shaft, Tunnel invert 500.99, POKFOOLUM ROAD, Waterworks Cottages and FARM LOT No. — a label that stops at No., with no figure after it. In red it records GAUGE BASIN across a two-chamber structure with a U-turn channel around it, NEW PATH TO GAUGE BASIN & … 6FT WIDE, the outlet works, and two levels, 499.77 and 501.35. Because the words Old and NEW are on the sheet, the reading that red marks the proposed works does not rest on drafting convention alone; convention merely confirms it. One distinction should be kept firmly in view: on the sheet itself the subject is labelled GAUGE BASIN in red, while New Gauge Basin appears in the title. The two must not be conflated when the label is quoted.

Tier 3 fixes the place and the datum. The new structure sits at the toe of the labelled EMBANKMENT, immediately below the impounding water body whose first letters survive as …RVOIR, and the tunnel that feeds it is given an invert level of 500.99 feet. Location and level are therefore stated on the artefact in a form that can be checked against documents without circularity.

Tier 4 supplies the outside corroboration, and it is deliberately drawn from four different series. Price’s report of 10 July 1875 proposes “a surface conduit of adequate dimensions, along the 500 foot contour from the Pokfoolum guage-basin to the Albany Tanks in Victoria.”6 The despatch forwarding the same scheme to the Colonial Office repeats it — “along the 500 foot contour from the Pokfulam storage basin to the highest Tanks in the city” — at the same figure of £9278.12.0.7 The Sessional Paper of 1896, looking back on the executed contract, says that the covered conduit began “at a Gauge basin to be constructed immediately below the Pokfoolum reservoir dam.”8 And The China Mail of 13 July 1878, describing the finished works, places the basin “at the foot of the outer slope of the Pokfoolum embankment, into which the water flows from the Reservoir and where the day’s supply is regulated and registered.”9 The Gazette text and the despatch are two records of one scheme by different authors and are counted as a single witness for that reason; the 1878 press account and the 1896 retrospective are independent of the planning papers and of each other, and both post-date the completed works.

Tier 5 is a numeric agreement. The sheet gives a tunnel invert of 500.99 feet; the documents give a “500 foot contour” and a level “of about 500 feet above the sea.” A drawing level and a textual level produced independently agree to within the precision of the drawing, and the basin levels 499.77 and 501.35 straddle the same datum. Tier 6 is provenance: the archive record and the artefact agree with each other on identity — original reference W00239 against the sheet’s own W 239 — and the file range 1871–1877 brackets the 1876 tender and the completion of 1877 without supplying either date directly.101112

What these six tiers establish is the subject, the location and the old/new relationship of the two basins. What they do not establish is set out in § 10: the sheet’s date, its author, any membership of the “twenty sheets of detailed drawings and estimates” that Price transmitted in 1875, the identity of the Old Gauge Basin, and the legal renaming of the alignment.13 One negative result belongs here because it shaped the whole method: an exact-phrase search for New Gauge Basin returns nothing at all in the corpus, so no contemporary document names this drawing.14 The identification therefore had to rest on the artefact’s own text plus four descriptions of the structure, never on a record pointing at the sheet.

3. Short supply and Price’s subsidiary works, 1873–1875

The conduit was designed into an emergency. From the end of 1863 the city had been fed by a single ten-inch cast-iron main laid under the Pokfulam Road carriage road, and by the mid-1870s the shortage was a standing item of correspondence. In February 1874 the newly appointed Surveyor-General, J. M. Price, wrote that “the supply, calculated at the usual rate of five gallons a head, will fall short by about twenty days”; in September of the same year a typhoon choked the existing works, for “the gauge basin and outlet pipe were so choked with ballast and sand that the water supply to the City had to be cut off for a day.”1516 A works vote of the same year provided for “re-ballasting Embankment, repairing and chunamming floor of Gauge basin and dam” — maintenance on the pre-existing structure that the sheet under study would later label, in the plainest possible terms, Old.17

Price’s answer was published as a report to the Legislative Council in the Hongkong Government Gazette of 10 July 1875. He transmitted the surveys “embodied in the accompanying twenty sheets of detailed drawings and estimates,” which included two alternative Tytam projects, the High Level Project at £39,000 and the Low Level Project at £122,600.18 Among them a separate section, headed Pokfoolum Subsidiary Works, proposed the conduit itself: a surface conduit on the 500-foot contour from the Pokfoolum gauge basin to the Albany Tanks, together with an iron main and a heading at the reservoir end, at a total estimate of £9,278. 12. 0 (conduit £6,864. 11. 0; iron main £1,297. 18. 0; heading at the reservoir end £272. 18. 0; contingencies at ten per cent).19

Two sentences in that report govern everything that can honestly be said about the works’ effect. The first concedes that “a new conduit could not increase the supply during that season”; the second is the design response — because the reservoir, left uncontrolled, could be emptied “in about six weeks,” the yield was to be rationed “by the insertion of a meter at the outlet works.”20 Measurement was built into the scheme from the first page, and that is why the gauge basin of the drawing is not a decorative appendage but the operational centre of the design. A caution attaches to the report’s estimate tables: the line items Culvert to Gauge Basin, Gauge Basin and Tunnel Inlet, Valve Shaft, Tunnel and Tunnel Outlet printed on the following page belong to the two alternative Tytam schemes, not to the Pokfoolum works, which are costed separately.21 Reading them across into this sheet would be exactly the kind of quiet conflation the evidence register exists to block.

The despatch that carried the scheme to the Colonial Office preserved both the figure and the caveat: Price proposed to “lay down a conduit along the 500 foot contour from the Pokfulam storage basin to the highest Tanks in the city…the cost is stated to be £9278.12.0,” while stating in the same file that “the proposed Conduit will in no way increase the daily quantity of water at present delivered during the dry season.”22 23 The construction specification followed on the next page: a stone bottom and brick sides rendered in neat Portland cement, a cover of rough granite slabs, cast-iron syphons across the ravines, three and a half miles in all, on a resulting gradient of 1 in 326.24 The old ten-inch main was not to be discarded but re-used: “when the Conduit is completed and in use it is proposed to take the present 10 inch Conduit … and to utilize it in the City distributing main.”25 By 1877 the Colonial Office could record that the subsidiary scheme had “finally been adopted, and the further work is now being actively carried out.”26

Price had also written to the Government in August 1875, five weeks after the report, in terms that state the logic of the thing exactly: “during the rainy months of the year the great advantage of the conduit would be apparent in bringing into the town a supply practically unlimited.”27 The word doing the work is rainy.

4. From paper to ground: tender and construction, 1876–1877

The scheme reached the ground in 1876. An item in the Executive Council minutes reads “Lok foolam Conduit. Permission to the Purveyor General to invite tenders for proceeding with the work of enlarging present” — an OCR fragment that is not dated on the page and is not dated here.28 The public act followed: Gazette notice No. 98 of 13 May 1876 states that “tenders will be received at this Office, until Noon of the 5th proximo, for works in connection with the Pokfulum Reservoir. Particulars may be had on application at the Surveyor General’s Office” — that is, closing at noon on 5 June 1876.29 The Sessional Paper of 1896 confirms the sequence in its own words: construction “commenced till 1876 when contracts were entered into for the construction of a covered conduit” from a gauge basin below the reservoir dam to the Albany Tanks.30

What was contracted is known in detail from the same retrospective. The conduit ran 17,840 feet in a brick and masonry rectangular section one foot six inches by one foot six inches; five cast-iron inverted syphons, two ten-inch pipes side by side, totalled 2,206 feet across the ravines; the discharging capacity was 2,000,000 gallons a day, falling to 1,700,000 beyond the fourth syphon. “The work was completed in 1877, and the 10-inch main in the Pokfulam Road removed at a total cost of $62,090.76, having been carried out under the supervision of Mr. W. Danby, C.E.”31 The press figure of July 1878 — “about sixty-two thousand dollars” — is an independent witness to the same sum.32 The mild conflict over completion, between the Sessional Paper’s 1877 and a newspaper sentence about works “now rapidly approaching completion” in July 1878, is resolved in favour of 1877 in the register rather than by fiat.33

The currency and scope limits here matter, and they are stated once. The £9,278. 12. 0 of 1875 is an estimate for the subsidiary works in pounds sterling; the $62,090.76 of 1877 is the actual cost of the whole conduit in dollars. They are not the same scope and not the same currency, and subtracting one from the other would produce a variance that never existed. A modern account that gives a completion cost of £13,038 is a single unreconciled web source and is not printed as a fact in this paper.34

One modern statement is worth having on the record because it agrees with the corpus on the numbers and can be checked against it: the Antiquities and Monuments Board’s consultation paper for N372 gives the same length, the same route from “the Gauge Basin just below the Pok Fu Lam Reservoir Embankment to the Albany Tanks,” the same cost of $62,090.76 and the same 1876–1877 chronology.35 It also supplies a superlative, which § 10 declines to adopt.

5. What was built: the conduit and its headworks

The sheet is not a map of the conduit. At twenty feet to the inch it is a large-scale detail of the works at the head of a 17,840-foot alignment: the reservoir outlet, the heading and tunnel with its valve shaft and an invert level of 500.99, the existing Old Gauge Basin and Old Dam, and the new gauge basin proposed in red with its outlet pipes and a new six-foot path. It is, precisely, the structure the Sessional Paper says the conduit “commenced at.”36 Recognising what kind of drawing it is — a headworks detail, not a route plan — prevents the two most obvious errors: treating it as evidence for the whole alignment, or looking on it for features that can only lie three and a half miles away.

The headworks themselves are described in the press account of July 1878, which has the water passing from the reservoir through the gauge basin and then through a tunnel driven “through the solid rock under the hill on which Douglas Castle stands,” after which it enters the conduit and runs to the Albany tanks along the contours of the hill at an elevation of 500 feet, for three and a half miles.37 The sheet shows exactly that arrangement — inlet, tunnel, valve shaft, invert 500.99, basin — and the two readings reinforce one another without either being derived from the other. What no document does is name the drawing and the villa’s hillside in one breath; the association is interpretation, well supported, and marked as such.38

The syphons were the vulnerable part. Two of them crossed the largest valleys as pairs of cast-iron pipes with granite well-heads, and their effect on discharge was not theoretical: the Public Works Department reported in 1913 that “the occurrence of syphons at two points on the route of the Pokfulam Conduit resulted in the discharging capacity of the conduit being seriously impaired and the removal of these syphons was therefore undertaken, lengths of conduit contouring the hillside being substituted for them.”39 Thirty-six years after completion the design was still being corrected on site, near Villa Lucia, and the correction took the same form the original scheme had taken: a contour line of masonry on the hillside.

6. What the conduit did and did not do

The honest account of the works begins with a concession made by their own designer. The conduit did not create water. Price wrote in 1875 that a new conduit “could not increase the supply during that season,” and the despatch to the Colonial Office repeated it almost verbatim: “the proposed Conduit will in no way increase the daily quantity of water at present delivered during the dry season.”4041 What the conduit did was capture in the wet months what would otherwise have gone over the bye-wash — water that the uncontrolled reservoir could not hold and that the city therefore never saw. The advantage was explicitly seasonal, and Price said so again in August 1875, when he told the Government that during the rainy months the conduit would bring in “a supply practically unlimited.”42

Measured against the old position the gain was nonetheless large, and it can be stated with its witness and its season attached. The ten-inch main had fed the city since the 1860s at something under five gallons a head a day — four and a half, in the press’s reckoning of 1878 — while the new arrangement was claimed in the same article to deliver twenty-two gallons per inhabitant per diem.43 Six years later the Governor put a figure of a different order before the Legislative Council: “the actual water supply consequent on the improvement in the Pokfulam conduit, the reservoir and filter beds” amounted to “an average quantity of 14½ gallons per head of inhabitants per day during the wet season and 15¾ gallons during the dry season.”44 That attribution is deliberately joint. The conduit, the reservoir and the filter beds are named together by the Governor, and no honest reading credits the conduit alone with the increase — a limit that survives into the companion post for this paper.

The context for the 1881 statement is worth having, because it shows how long the argument took. In 1878 the Governor reported that the reservoir “was only giving one-third the quantity of water which it had been constructed to supply”; Moorsom’s explanation was that the pipe had been too narrow and that “a large conduit was required”; and Price’s own letter of February 1874 was recalled in the same file, with the added remark that “at the date of that discussion this large conduit had not been constructed.”45 The conduit was, in other words, the answer to a problem that had been correctly stated four years before it was authorised.

A further limit is arithmetic rather than rhetorical. A modern history states that the reservoir’s dry-season abstraction averaged 474,218 gallons a day, a figure that belongs to Price’s letter of August 1875 and to the reservoir, not to the conduit’s carrying capacity of two million gallons a day.46 Per-head rates are quoted throughout this paper only with their witness and season attached; no total consumption or population figure is asserted anywhere, for the simple reason that the corpus does not support one.

7. The basin as instrument: measurement and regulation

If the conduit was a way of carrying water, the gauge basin was the way of knowing how much water there was. Price’s scheme of 1875 specified that the yield be rationed “by the insertion of a meter at the outlet works,” because a reservoir that could be emptied in six weeks could not be trusted.47 The 1878 description of the completed works shows the same idea at the other end of the pipe: the gauge basin lies “at the foot of the outer slope of the Pokfoolum embankment, into which the water flows from the Reservoir and where the day’s supply is regulated and registered.” 48 Register and regulate, not store: the structure drawn in red on the sheet is an instrument panel in masonry.

The system grew a second reading station within a few years. The China Mail’s companion column reports that above the town, where the supply branched off to the Robinson Road service tank, “a second tell-tale gauge basin has been constructed for the purpose of re-measurement, so as to ascertain the extent of loss of water in transit by leakage or absorption.”49 Two gauges, one at each end, with the difference between them standing for what the hillside had absorbed: measurement had become a way of auditing the conduit itself. The instrument remained a standard component of Hong Kong waterworks engineering thereafter. The 1907 estimates for the filter beds provide for “a new conduit, gauge basin, service and washout pipes between the filter beds and service reservoir, sand screening pit and eight Venturi meter chambers”;50 the 1917 works list has a “Recorder, Gauge Basin, Albany Filter Beds (not quite completed)”;51 and the 1924 ledger still carries the line “Unfiltered water, Gauge Basin, Albany Road.” In 1918 the same department built two gauge basins, one at the new filter beds and one at the Albany filter beds, “for measuring the water,” with a break-pressure tank on the supply from the Pokfulam Conduit.52 Forty years after the sheet was drawn, its central idea was being re-cut in concrete at other points in the system.

Measurement implies custody, and the records show custody in its plainest form. In 1891 the vacant post of “Watchman on the Pokfulam Conduit” was declined, the duties being merged with those of the watchman at the Pokfulam Filter Beds rather than allowed to lapse;53 the Public Works Return of January 1882 lists among the incidental services of maintenance “repairing, covering stones Pokfulam Conduit” five years after completion;54 and in 1904 the conduit was reported “thoroughly cleaned and freed from obstructions throughout its entire length,” with a vote for improvements alongside it.55 A covered conduit of flat gradient and loose stone slabs was a thing that had to be swept, watched and counted.

8. The conduit under the city

The alignment did not stay a waterworks feature for long; within a decade it had become a line in the urban fabric. The first sign is an 1884 vote of $950 “for resumption by the Crown of a portion of Inland Lots 714 and 715 on Robinson Road, on account of the Pokfulam Conduit being under it.”56 The Crown had to buy back ground above its own pipe: reticulation had been carried above Robinson Road for the first time, and the land above the conduit had begun to be built over the conduit.

The next sign is boundary language. The Public Works Department report for 1891 records that the training of the nullahs in the Western District had made “considerable progress … with the works of training the Nullahs in the Western District between the Pokfulam Conduit and the Robinson Road,” at a cost to date of $12,640.52.57 By 1891 the conduit was not merely a pipe; it was a boundary against which public works were described.

By then it had also become a zoning limit. Chadwick’s report of 28 December 1889 divided the city into three zones for sewerage, of which “the uppermost zone from the 250-feet contour to the Pokfoolum conduit” was the highest; the same paper observed that the built area had by then “extended in places up to and even above the level of the Pokfoolum conduit,” and the works were “finally decided upon and carried out during the years 1890-1892.”58 Read against the sheet, this is the trajectory of the works drawn in red: proposed below the dam in the mid-1870s, carrying water above the old town by the 1880s, and serving as the ceiling of the city’s highest sewerage zone by the end of the decade. The filter beds and service reservoir completed in 1889–90 sit at the same moment, and the corpus’s modern heritage papers place them there.59

9. From conduit to Conduit Road, 1899–1948

The last transformation is the one still visible from the street. On 4 October 1899 the Legislative Council referred a list of proposed works to the Public Works Committee, item (10) of which is “Pokfulam Conduit Road.”60 The Sessional Paper of 1900 explains why: “Owing to several sites for residences being sold above the level of the Robinson Road sites, it became necessary to construct a road giving access to them,” and a commencement had been made with the approach road up Glenealy Ravine; the annual report for the same year adds that “good progress was made on the Pokfulam Conduit Road during the year, and plans and estimates are ready for its completion as far as a junction with Robinson Road.”61 The works ran on: carried to Inland Lot 1549 during 1901 at a cost of $9,684.73;62 completed in November 1902 “with the exception of a short length near its western extremity,” and widened from twelve feet to sixteen;63 and addressed by residents as “Conduit Road” in the Directory by 1904.64 By 1918 an Executive Council minute could take it for granted, describing acquired property as Inland Lot 1546 “known as No. 31, Conduit Road” and Inland Lot 1548 as “Nos. 17, 19, 21, 23, 25, and 27, Conduit Road.”65 The alignment of the sheet’s “NEW PATH” and outlet pipes had become an address. The gazetted street name in modern Chinese is 干德道, checked against the Lands Department list rather than reconstructed.66

The old works continued in service under the road. The Gazette’s list of roads closed in August 1914 includes, as item 1, “Pokfulam Conduit west of Victoria Battery” — the conduit walk itself treated as a road closed to the public.67 Landslips in the 1906 typhoon “threatened but did not seriously injure the Pokfulam Conduit”;68 two two-inch mains were taken off it in 1922 to serve houses on New Mount Davis Road and in the Villa Miramere district;69 and as late as the Tai Lam Chung engineering paper of 1948 the conduit stood first in the list of the colony’s four trunk mains, with the frank assessment that “there is no doubt that the brick conduits have given excellent service, but their capacity is limited to the free discharge on very flat gradients, for being covered only with loose stone slabs they do not permit of any pressure head.”70 The same paper states that the four trunk mains were damaged “one after the other” in the 1941 attack without saying which, when or how badly; no damage to this conduit is established, and none is claimed.71

Heritage recognition came more than a century after the sheet was drawn: the No. 9 aqueduct was graded Grade 2 in 2004, and the conduit was the subject of a public consultation by the Antiquities and Monuments Board on its status and improvement.72 That modern body is also the source of the superlative discussed in § 10, and of the 1891 Gazette notification this research was unable to find.73

10. Evidence register: conflicts, gaps and do-nots

What follows is the part of the method that is usually invisible. Every item below was either tested and passed, tested and downgraded, or tested and failed; the register is written before the companion post so that nothing enters publication that has not already been adjudicated here.

10.1 Confirmed: two or more independent contemporary witnesses

Items resting on two or more independent documents.
ItemStatusWitnesses
Title, scale (20 ft = 1 in), sheet mark W 239, no date on the sheetConfirmedArtefact read at enlargement; PRO catalogue record
Conduit let in 1876 from a gauge basin immediately below the reservoir dam, contouring at about 500 feet to the Albany TanksConfirmedSessional Paper 1896; Gazette tender No. 98, 13 May 1876
17,840 ft; 1′6″ × 1′6″; five syphons totalling 2,206 ft; 2,000,000 → 1,700,000 gal/dayConfirmedSessional Paper 1896 (official retrospective)
Completed 1877; 10-inch main removed; $62,090.76; supervised by W. Danby, C.E.ConfirmedSessional Paper 1896; China Mail, 13 July 1878 (“about sixty-two thousand dollars”)
Estimate £9,278. 12. 0 for the Pokfoolum subsidiary worksConfirmedPrice’s report (Gazette, 10 July 1875); CO129-171 — agree to the penny
Conduit laid on the 500-foot contour from the gauge basin to the Albany TanksConfirmedGazette 1875; CO129-171; China Mail 1878; Sessional Paper 1896
The gauge basin as the measuring and regulating pointConfirmedPrice’s meter at the outlet works (1875); “regulated and registered” (1878); filter-bed gauge basins (1917–1918); Venturi chambers (1907)
The conduit did not increase dry-season yieldConfirmedGazette 10 July 1875; CO129-171, almost in the same words
Delivery above Robinson Road; 1884 resumption of I.L. 714–715 because the conduit ran under themConfirmedSessional Paper 1884; AMO board paper
A road on the conduit alignment, “Pokfulam Conduit Road” (1899) and later addressed “Conduit Road”ConfirmedGazette L.C. minutes 1899; Sessional Papers 1900–1902; CO131 1918; Directory 1904
Syphons removed 1912–13, contouring lengths substitutedConfirmedPublic Works Department Annual Report 1913

10.2 Single-witness, interpretive, or reconstruction notes

The sheet’s date remains an inference. It is probably 1875–76 — after Price’s report of 10 July 1875, which is the earliest contemporary description of the scheme, and before or around the tender of May 1876 — because the sheet shows the new basin as proposed and refers to an existing old basin and old dam, both of which are attested in 1874–75 records. But no date is printed on it, and 1871–1877 is a file range. This paper therefore prints no bare year for the sheet as a fact.7475 The tunnel on the sheet is read as the tunnel the press described through the rock below Douglas Castle: the arrangement matches and no document contradicts it, but no document names sheet and hillside together either.76 The press quotations of July 1878 are reconstructed from column-interleaved OCR and are quoted as the newspaper’s own summary, not as a specification.77 The red label on the channel is probably OUTLET PIPES FROM RESERVOIR, a probable reading from a later enhancement pass; it is not printed here as a quotation, and neither are the stray red letters T N N E over the farm lot. The identification of the Old Gauge Basin with Governor MacDonnell’s 1866 proposal of a basin holding 300,000 gallons rests on a single modern account and no primary source ties it to the sheet’s label.78 The Executive Council minute is not dated here. The 1881 per-head figures are jointly attributable and are not credited to the conduit alone.79 The completion date stands at 1877 despite the 1878 press line.80

10.3 Failed verification and corrections

Claims tested and not carried.
Claim or leadResult
Any document that names the drawingZero hits. Exact phrase “New Gauge Basin” returns 0 in hkintexts_pages (6 October 2026); “Gauge Basin” returns 7, none an 1870s reference to a Pokfoolum drawing.81
AMO citation of Notification No. 490, Gazette, 28 November 1891, for the conversion into a roadNot found. “Notification No. 490” returns 16 hits, none from 1891 and none about a road; a full-phrase run over all keys returns no 1891 Gazette document key at all. The 7 October 1899 element of the same note is corroborated.82
A remembered citation of a 1910 PWD report (https://hkintexts.histsyn.com/page/613600; accessed 5–6 October 2026) describing a gauge basinCorrected. Re-fetched 5 October 2026: the id is a rainfall paragraph. Dropped.83
Four year-sliced search labels (pf1875–pf1878)Invalid runs. Issued without a document_key filter, so all four returned the same unfiltered total. No year-sliced count is quoted.84
Damage to the conduit in the 1941 attackNot established. The 1948 paper says the four trunk mains were damaged “one after the other” without naming which.85
Signature or authorship on the sheetNot found. Bottom-left, bottom-centre and title-block regions examined; no signature (contrast the Taipingshan plan, signed F. A. Cooper).
An earlier reading of the sheet label as “FARM LOT No 8”Corrected 6 October 2026. A tight re-read at 8–14× magnification with contrast stretch shows the label ending at No., with nothing after it before the paper loss. No lot number is printed anywhere in this paper, and the corpus hit land-9070 / Farm-Lot/8 is not attributable to the sheet: the number came from the search, not from the artefact.

10.4 Negative controls: what must not be conflated

Two controls guard the identification against cross-contamination. The first is the Tytam gauge basin: the description at CO129-171, p. 300, of water “passed through a small quage basin furnished with a proper guage” before entering a conduit 1,356 yards long belongs to the Low Level (Tytam) Project altogether, and is cited here only so that it can never be mistaken for this sheet’s subject.86 The second is the estimate tables of the 1875 report itself: the Gauge Basin line items on printed p. 283 belong to the High Level and Low Level Tytam estimates, while the Pokfoolum subsidiary works are costed separately.87

10.5 Do-nots

Twelve prohibitions are carried from the register into every derived text. Do not print a bare year for the sheet; do not describe the drawing as the conduit itself (it is a headworks detail of a 17,840-foot alignment); do not claim it is one of Price’s twenty sheets;88 do not claim the conduit increased the dry-season supply; do not credit the 1881 figures to the conduit alone; do not use “first,” “oldest,” “earliest” or “largest” — including the Antiquities and Monuments Board’s own “earliest masonry and brickwork conduit in Hong Kong,” which has no second independent primary witness and is therefore not repeated as a finding;89 do not conflate the Pokfulam gauge basin with the Tytam gauge basins; do not reconstruct the illegible red label or the stray letters; do not print a lot number for the sheet’s farm-lot label, which stops at No.; do not assert a date for the Executive Council minute; do not subtract the dollar actual from the pound estimate; and quote no total population, consumption or death figure — only per-head rates with witness and season attached.90

11. Conclusion

A drawing without a date can still be identified, provided the identification is built in layers and every layer is allowed to fail on its own. The sheet names itself; its labels separate Old from NEW in words; it places the new basin at the toe of the embankment with an invert of 500.99 feet; four documents in four series describe a gauge basin at exactly that place; a drawing level agrees with their contour to within a foot; and the archive record and the pasted label agree on the number. That is enough to say what the sheet is, and it is not enough to say when it was drawn, who drew it, or whether it was one of the twenty sheets Price sent to the Legislative Council in July 1875. The second half of the identification — the part that is not established — is published alongside the first, because a paper that cannot state its own limits cannot be checked by anyone else.

The works themselves repay the attention. A conduit of 17,840 feet, five syphons and two million gallons of daily capacity was built not to make water but to carry water that the hills had already given, and its designers said so in the same breath as they priced it. The basin at its head existed to turn flow into a number: “the day’s supply is regulated and registered,” a second tell-tale basin above town to audit the loss in transit, and by the 1900s Venturi chambers and recorders at the filter beds. The conduit then did what successful infrastructure does and disappeared into the city: a boundary for nullah training in 1891, the ceiling of Chadwick’s uppermost sewerage zone, the reason for a Crown resumption on Robinson Road in 1884, and by 1902 a road widened from twelve to sixteen feet with house numbers by 1918. Today it is 干德道, Conduit Road — a name that still records what runs underneath it, and the reason the sheet in the Public Records Office is worth reading at twenty feet to the inch.

What this paper will not claim is as much a part of it as what it does. The drawing is not the conduit; the conduit was not an increase in dry-season supply; the sheet is not proven to be one of Price’s twenty; no document in the corpus names it; the 1891 Gazette notification could not be found; and no “earliest ” attaches to it. Every one of these limits travels into the Cantonese post accompanying this paper, which is written to say the same things in fewer words.91

Appendix A: ES query log

All queries were run read-only against the hkintexts ES cluster, index hkintexts_pages, on 5–6 October 2026, with tools/search.py, tools/show.py, tools/dumpdoc.py and tools/get.py. The log below is representative; full payloads are retained with the working papers and every corpus citation in the endnotes and the bibliography carries its public viewer link (https://hkintexts.histsyn.com/page/{id}) with an access date.

Representative searches; hits are as returned on the dates given.
LabelQuery (phrase / filter)IndexHits
pfc1876match_phrase “Pokfulam Conduit”hkintexts_pages171
s_gauge_basinmatch_phrase “gauge basin”hkintexts_pages7
s_New_Gauge_Basinmatch_phrase “New Gauge Basin”hkintexts_pages0
s_Pokfulam_Conduit_Roadmatch_phrase “Pokfulam Conduit Road”hkintexts_pages22
gb76match_phrase “Gauge Basin” + regexp .*187[4-9].*hkintexts_pages4
pok76match_phrase “Pokfulum Reservoir” + regexp .*187[4-9].*hkintexts_pages7
(inline)match_phrase “Lok foolam Conduit” + prefix page.1874-1875-1876-co131-9hkintexts_pages1
(inline)match_phrase “Pokfulam Conduit Road” + regexp .*(189[0-9]|190[0-5]).*hkintexts_pages37
(inline)match_phrase “Conduit Road” + regexp Directory_and_Chronicle_190*hkintexts_pages47
(inline)match_phrase “Pokfulum” + regexp Directory_and_Chronicle_189*hkintexts_pages0
n1891_conduitmatch_phrase “Pokfulam Conduit” (size 100), grouped by document_keyhkintexts_pages171
no 1891 Gazette key
n1891_notif490match_phrase “Notification No. 490”hkintexts_pages16
land_farm8lot_label="Farm Lot" AND lot_number_num=8hkintexts_land_lots1
not attributable
id re-verification{"ids":{"values":[36 ids]}}, comparing id, document_key, page_numberhkintexts_pages36/36
matched
new-id fetchestools/get.py for 6 later-added ids (617399, 611000, 1815313, 347224, 1665232, 584000)hkintexts_pages6/6
read

Invalid runs, not cited: labels pf1875, pf1876, pf1877, pf1878 were issued without a document_key filter and all returned the identical unfiltered total of 5,614, overwriting the same results file. Any year-sliced count must be re-run with a regexp or prefix filter first.92

Other conventions: no writes were made to the cluster; the miniature replica at 100.105.69.105:9200 was not used; the image-page index hkintexts_image_pages returns no hits for phrase queries of this kind and its silence was not read as absence of evidence; and where OCR of a newspaper page is column-interleaved the sentence was reassembled and read on the full page before being quoted.93

Notes

  1. ↑Government Records Service (Hong Kong), catalogue record HKRS1951-1-3, Hong-kong Waterworks – Pokfoolum Conduit. New Gauge Basin, series HKRS1951 Drawings Relating to Water Works, original ref W00239, file dates 1871–1877, access under the Public Records (Access) Rules 1996, https://search.grs.gov.hk/en/arcview.xhtml?q=HKRS1951-1-3&eid=hlz6QKooYl3kDojwfO8Erw%3D%3D&ls=e_k%3D1%26q%3DHKRS1951-1-3%26rpp%3D10 (accessed 5 October 2026). The 1871–1877 range is a file range, not a date on the sheet.
  2. ↑Method and corpus. The paper is artefact-led: the drawing was read first (controlled crops, contrast and red-channel enhancement, rotation), and only then was the corpus queried. The corpus is the hkintexts ES cluster (read-only; index hkintexts_pages), interrogated 5–6 October 2026. Twenty logged searches and thirty-six bulk id re-verifications are itemised in Appendix A. Every quotation below was read on a page pulled in full with tools/dumpdoc.py or tools/get.py, not on a search-result snippet, and every document id cited in the endnotes was re-fetched through its public viewer link and matched against its printed page before it was printed.
  3. ↑Corpus caveat. The Chinese-language press in this index begins about 1926 (daily papers about 1947), so the 1870s stratum is English-only; a silence in this paper about Chinese-language reporting is a silence of the corpus, not evidence of historical absence. Period place spellings are kept as printed (Pokfoolum, Pokfulum, Pokfulam, Pok Fu Lam). Modern Chinese forms were checked rather than reconstructed: Conduit Road = 干德道 and Albany Road = 雅賓利道 in the Lands Department gazetted street-name list; the Public Records Office = 香港政府檔案處 (grs.gov.hk).
  4. ↑[Artefact] HONG-KONG WATERWORKS. POKFOOLUM CONDUIT. NEW GAUGE BASIN., ink and wash on paper, 20 feet = 1 inch, sheet mark “W 239” (pasted catalogue label), red docket reading “⁆/2. DRAW.  No. 1.”, no printed date and no signature. Digital file Pokfoolum Conduit/HKRS1951-1-3_0001.jpg (3043 × 2048 px). Read 5–6 October 2026. On-sheet labels quoted in this paper are transcribed as they stand: Old Gauge Basin, Old Dam., EMBANKMENT, Valve Shaft, Tunnel invert 500.99, GAUGE BASIN (red), NEW PATH TO GAUGE BASIN & … 6FT WIDE (red), POKFOOLUM ROAD, FARM LOT No. (no figure follows No.), Waterworks Cottages.
  5. ↑Government Records Service (Hong Kong), catalogue record HKRS1951-1-3 (file dates 1871–1877; accessed 5 October 2026).
  6. ↑Price, “Report on the Water Supply of Hongkong,” Gazette, 10 July 1875, 284, section “Pokfoolum Subsidiary Works”: “lay down a surface conduit of adequate dimensions, along the 500 foot contour from the Pokfoolum guage-basin to the Albany Tanks in Victoria.” https://hkintexts.histsyn.com/page/1673081 (accessed 5–6 October 2026).
  7. ↑Colonial Office, Hong Kong correspondence series CO129, CO129-171 – Acting Governor Austin – 1875 [7-11] – Sir Kennedy – 1875 [12], 304: “It is now proposed by W. Price to lay down a conduit along the 500 foot contour from the Pokfulam storage basin to the highest Tanks in the city … the cost is stated to be £9278.12.0.” https://hkintexts.histsyn.com/page/296574 (accessed 5–6 October 2026).
  8. ↑Sessional Papers, Hong Kong, 1896, paras 17–19: construction “commenced till 1876 when contracts were entered into for the construction of a covered conduit commencing at a Gauge basin to be constructed immediately below the Pokfoolum reservoir dam and contouring the hillside at a level of about 500 feet above the sea to the Albany Tanks.” https://hkintexts.histsyn.com/page/1818318 (accessed 5–6 October 2026).
  9. ↑“The New Pokfoolum Waterworks,” The China Mail, 13 July 1878, 6: the gauge basin lies “at the foot of the outer slope of the Pokfoolum embankment, into which the water flows from the Reservoir and where the day’s supply is regulated and registered”; the tunnel runs “through the solid rock under the hill on which Douglas Castle stands”; the conduit “extends to the City so far as the Albany tanks, winding along the contours of the hill at an elevation of 500 feet above the sea for the distance of three and-a-half miles”; the old supply pipe “which has fed the city since 1862” gave “four and a half gallons per head” against “twenty-two gallons per inhabitant per diem” for the new conduit. https://hkintexts.histsyn.com/page/1890648 (accessed 5–6 October 2026). The page is OCR’d as interleaved columns; each sentence was reassembled from the fragments and read on the full page, and is quoted as China Mail’s own summary rather than as an engineering specification.
  10. ↑Government Records Service (Hong Kong), catalogue record HKRS1951-1-3 (file dates 1871–1877; accessed 5 October 2026).
  11. ↑Gazette, No. 98, 13 May 1876, 235 (https://hkintexts.histsyn.com/page/1672064; accessed 5–6 October 2026).
  12. ↑Sessional Papers, 1896, 252 (printed 249; https://hkintexts.histsyn.com/page/1818319; accessed 5–6 October 2026).
  13. ↑On what is claimed. The identification offered in § 2 establishes the subject, the location and the old/new relation of the two basins. It does not establish the sheet’s date, its authorship, its membership in Price’s twenty sheets, the identity of the “Old Gauge Basin” or the legal renaming of the conduit alignment. Those limits are carried into the Cantonese post companion to this paper; nothing here is offered as a first, an earliest or a longest.
  14. ↑On negative results. An exact-phrase search for "New Gauge Basin" in hkintexts_pages returns zero hits (label ngb, 6 October 2026), and "Gauge Basin" returns seven, none of them an 1870s reference to a Pokfulam drawing. No contemporary document, therefore, names this sheet; the identification in § 2 rests on the artefact’s own text and on four descriptions of the structure in four different series. A second negative: searching "Pokfulam Conduit" over all keys returns 171 pages and no 1891 Gazette document key at all, so the Gazette notification of 28 November 1891 cited by the Antiquities and Monuments Board for the road conversion could not be corroborated ("Notification No. 490" returns sixteen hits, none from 1891 and none about a road).
  15. ↑CO129-193, 164: Kennedy’s 1878 report that the reservoir “was only giving one-third the quantity of water which it had been constructed to supply”; Moorsom’s explanation that “a large conduit was required”; Price’s letter of 18 February 1874, “The supply, calculated at the usual rate of five gallons a head, will fall short by about twenty days”; and the remark that “at the date of that discussion this large conduit had not been constructed.” https://hkintexts.histsyn.com/page/308824 (accessed 5–6 October 2026).
  16. ↑Colonial Office, CO129-168 – Sir Kennedy – 1874 [9-12], 220: after the September 1874 typhoon “The gauge basin and outlet pipe were so choked with ballast and sand that the water supply to the City had to be cut off for a day to allow the outlet works to be cleaned and repaired.” https://hkintexts.histsyn.com/page/294250 (accessed 5–6 October 2026). This is the pre-existing works labelled on the sheet as Old Gauge Basin.
  17. ↑Colonial Office, CO129-168, 301: a works vote includes “re-ballasting Embankment, repairing and chunamming floor of Gauge basin and dam, purchasing boat.” https://hkintexts.histsyn.com/page/294331 (accessed 5–6 October 2026).
  18. ↑J. M. Price, Surveyor-General, “Report on the Water Supply of Hongkong,” Hongkong Government Gazette, 10 July 1875, 282 (Price transmits the surveys “embodied in the accompanying twenty sheets of detailed drawings and estimates”), https://sunzi.lib.hku.hk/hkgro/view/g1875/694474.pdf (accessed 5 October 2026). https://hkintexts.histsyn.com/page/1673079 (accessed 5–6 October 2026). The report never lists the twenty sheets, so no sheet can be tied to this set; see the do-not in § 10.
  19. ↑Same, 285: the estimate for the subsidiary works is “£9,278. 12. 0” (conduit £6,864. 11. 0; iron main £1,297. 18. 0; “Heading at Reservoir end” £272. 18. 0; contingencies 10 per cent £843. 5. 0); Price adds that “a new conduit could not increase the supply during that season” and that the yield must be rationed “by the insertion of a meter at the outlet works.” The reservoir, uncontrolled, could be emptied “in about six weeks.” https://hkintexts.histsyn.com/page/1673082 (accessed 5–6 October 2026).
  20. ↑Price, “Report on the Water Supply,” Gazette, 10 July 1875, 285 (https://hkintexts.histsyn.com/page/1673082; accessed 5–6 October 2026).
  21. ↑Same, 283: the estimate tables on this page belong to the two alternative Tytam schemes — the High Level Project (£39,000) and the Low Level Project (£122,600) — and it is their line items that read Culvert to Gauge Basin, Gauge Basin and Tunnel Inlet, Overflow Weir and Bye-wash, Valve Shaft, Tunnel and Tunnel Outlet. https://hkintexts.histsyn.com/page/1673080 (accessed 5–6 October 2026). Those basin line items are a negative control and must not be read as the structures drawn on the sheet under study; the Pokfoolum subsidiary works are costed separately.
  22. ↑CO129-171, 304 (https://hkintexts.histsyn.com/page/296574; accessed 5–6 October 2026).
  23. ↑CO129-171, 307: “the proposed Conduit will in no way increase the daily quantity of water at present delivered during the dry season.” https://hkintexts.histsyn.com/page/296577 (accessed 5–6 October 2026).
  24. ↑CO129-171, 308: the advantage of the conduit appears “during the rainy months of the year when the reserving is overflowing”; construction is specified as a stone bottom and brick sides rendered in neat Portland cement with a rough granite-slab cover, cast-iron syphons across the ravines, a total length of three and a half miles and a resulting gradient of 1 in 326. https://hkintexts.histsyn.com/page/296578 (accessed 5–6 October 2026).
  25. ↑CO129-171, 309: “When the Conduit is completed and in use it is proposed to take the present 10 inch Conduit … and to utilize it in the City distributing main.” https://hkintexts.histsyn.com/page/296579 (accessed 5–6 October 2026).
  26. ↑Colonial Office, CO129-177 – Sir Kennedy Acting Governor Austin Lieut Governor Hennessy – 1877 [1-5], 156: the Surveyor-General “submitted schemes, with a subsidiary proposal for a new conduit from the Pokfulum Reservoir … have finally been adopted, and the further work is now being actively carried out.” https://hkintexts.histsyn.com/page/299826 (accessed 5–6 October 2026).
  27. ↑CO129-193, 165: Price’s letter of 27 August 1875, “During the rainy months of the year the great advantage of the conduit would be apparent in bringing into the town a supply practically unlimited”; dry-season abstraction from the reservoir averaged 474,218 gallons a day. https://hkintexts.histsyn.com/page/308825 (accessed 5–6 October 2026).
  28. ↑Executive Council minutes, CO131-9 (1874 1875 1876), 286, item 20: “Lok foolam Conduit. Permission to the Purveyor General to invite tenders for proceeding with the work of enlarging present” (the OCR is fragmentary). https://hkintexts.histsyn.com/page/5221607 (accessed 5–6 October 2026). The minute is not dated here: it sits between a page referring to 11 May 1876 and a page signed in a name that does not fit 1876, so no date is asserted for it.
  29. ↑Hongkong Government Gazette, No. 98, 13 May 1876, 235: “Tenders will be received at this Office, until Noon of the 5th proximo, for works in connection with the Pokfulum Reservoir. Particulars may be had on application at the Surveyor General’s Office.” The closing date is therefore 5 June 1876. https://sunzi.lib.hku.hk/hkgro/view/g1876/692299.pdf (accessed 5 October 2026). https://hkintexts.histsyn.com/page/1672064 (accessed 5–6 October 2026).
  30. ↑Sessional Papers, 1896, 251 (printed 248; https://hkintexts.histsyn.com/page/1818318; accessed 5–6 October 2026).
  31. ↑Sessional Papers, Hong Kong, 1896, paras 20–21: total length 17,840 feet; brick and masonry rectangular section 1 ft 6 in × 1 ft 6 in; five cast-iron inverted syphons totalling 2,206 feet; discharging capacity 2,000,000 gallons a day, reduced to 1,700,000 beyond syphon no. 4; “The work was completed in 1877, and the 10-inch main in the Pokfulam Road removed at a total cost of $62,090.76, having been carried out under the supervision of Mr. W. Danby, C.E.” https://hkintexts.histsyn.com/page/1818319 (accessed 5–6 October 2026).
  32. ↑The China Mail, 13 July 1878, 3: the iron syphons cross two of the large valleys; “At a spot above the town where a portion of the supply branches off into the large Service tank on the Robinson Road, a second tell-tale gauge basin has been constructed for the purpose of re-measurement, so as to ascertain the extent of loss of water in transit by leakage or absorption”; total cost “about sixty-two thousand dollars”; works “carried out under the personal supervision of Mr William Danby, as executive engineer.” https://hkintexts.histsyn.com/page/1890645 (accessed 5–6 October 2026). Same column-interleaved caveat.
  33. ↑On the completion date. Sessional Papers, 1896, states that “The work was completed in 1877” (https://hkintexts.histsyn.com/page/1818319; accessed 5–6 October 2026), while The China Mail of 13 July 1878 opens with particulars of works “now rapidly approaching completion” (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026). The newspaper sentence is undated in its referent and may concern the filter beds or the syphons rather than the conduit. The Sessional Paper is kept; the press line does not move the date.
  34. ↑Industrial History Hong Kong, “Hong Kong Water Supply – Pok Fu Lam Reservoir,” https://industrialhistoryhk.org/hong-kong-water-supply-pok-fu-lam-reservoir/ (accessed 5 October 2026): the estimate breakdown (conduit £6,863; iron main £1,297; heading at the reservoir end £272; total £9,278), the rock tunnel “from the gauging basin through the adjacent hillside”, and the statement that the work “was completed … in 1877 at a cost of £13,038.” The £13,038 figure is a single modern source, is not reconciled with the contemporary $62,090.76, and is not printed in this paper.
  35. ↑Antiquities and Monuments Board (Hong Kong), N372 Pok Fu Lam Conduit Public Consultation (board paper), https://www.aab.gov.hk/filemanager/aab/en/content_4/N372_Public%20Consultation%20(ENG).pdf (accessed 5 October 2026): the conduit “was constructed by the Government in 1877”; construction “began in 1876”; “17,840 feet (around 5.4 km) from the Gauge Basin just below the Pok Fu Lam Reservoir Embankment to the Albany Tanks on the eastern side of the Glenealy Ravine … at a cost of $62,090.76”; n. 13 states that “In 1899, the Government agreed to convert this section into ‘Pokfulam Conduit Road’,” citing Notification No. 490, Gazette, 28 November 1891 and Notification No. 50, Gazette, 7 October 1899. Its superlative “the earliest masonry and brickwork conduit in Hong Kong” is a do-not (see § 10), and its 1891 notification could not be found in the corpus.
  36. ↑Sessional Papers, 1896, 251 (printed 248; https://hkintexts.histsyn.com/page/1818318; accessed 5–6 October 2026).
  37. ↑China Mail, 13 July 1878, 6 (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026).
  38. ↑Gwulo (Old Hong Kong), “The Pokfulam Conduit,” 11 March 2011, https://gwulo.com/pokfulam-conduit (accessed 5 October 2026): the conduit “commenced in the gauge basin just below the old dam (old pokfulam reservoir), followed the contour along the hillside about 500 feet above the sea, and ended at the Albany Service reservoir”; “the path of the old conduit above mid-levels is immortalised by the name of the road that was built on top of it, Conduit Road.” The same page is the single modern witness for the 1866 MacDonnell proposal of a 50-foot embankment whose backing would form a gauge basin of 300,000 gallons; no primary source ties that proposal to the sheet’s “Old Gauge Basin”.
  39. ↑Public Works Department Annual Report, 1913, 76: “(ix.) Clearing Pokfulam Conduit near Villa Lucia … The occurrence of syphons at two points on the route of the Pokfulam Conduit resulted in the discharging capacity of the conduit being seriously impaired and the removal of these syphons was therefore undertaken, lengths of conduit contouring the hillside being substituted for them.” https://digitalrepository.lib.hku.hk/catalog/05747t454 (accessed 5 October 2026). https://hkintexts.histsyn.com/page/198872 (accessed 5–6 October 2026).
  40. ↑Price, “Report on the Water Supply,” Gazette, 10 July 1875, 285 (https://hkintexts.histsyn.com/page/1673082; accessed 5–6 October 2026).
  41. ↑CO129-171, 307 (https://hkintexts.histsyn.com/page/296577; accessed 5–6 October 2026).
  42. ↑CO129-193, 165 (https://hkintexts.histsyn.com/page/308825; accessed 5–6 October 2026).
  43. ↑China Mail, 13 July 1878, 6 (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026).
  44. ↑Colonial Office, CO129-193 – Governor Hennessy – 1881 [5-7], 163: the Governor tells the Council of “the actual water supply consequent on the improvement in the Pokfulam conduit, the reservoir and filter beds” as “an average quantity of 14½ gallons per head of inhabitants per day during the wet season and 15¾ gallons during the dry season.” https://hkintexts.histsyn.com/page/308823 (accessed 5–6 October 2026). The attribution is explicitly joint; the conduit may not be credited alone.
  45. ↑CO129-193, 164 (https://hkintexts.histsyn.com/page/308824; accessed 5–6 October 2026).
  46. ↑CO129-193, 165 (https://hkintexts.histsyn.com/page/308825; accessed 5–6 October 2026).
  47. ↑Price, “Report on the Water Supply,” Gazette, 10 July 1875, 285 (https://hkintexts.histsyn.com/page/1673082; accessed 5–6 October 2026).
  48. ↑China Mail, 13 July 1878, 6 (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026).
  49. ↑China Mail, 13 July 1878, 3 (https://hkintexts.histsyn.com/page/1890645; accessed 5–6 October 2026).
  50. ↑Sessional Papers, Hong Kong, 1907, 1106: “in addition to the above, a new conduit, gauge basin, service and washout pipes between the filter beds and service reservoir, sand screening pit and eight Venturi meter chambers are included in the contract.” https://hkintexts.histsyn.com/page/1827294 (accessed 5–6 October 2026).
  51. ↑Administrative Reports, Hong Kong, Annual Report 1918, 440: “in order to admit of unfiltered water from Pokfulam Reservoir being used, when available, a connection with the Pokfulam Conduit was also provided. The work included the construction of two gauge basins,—one at the new Filter Beds and one at the Albany Filter Beds,—for measuring the water, a break-pressure tank on the supply from Pokfulam Conduit…” https://hkintexts.histsyn.com/page/617906 (accessed 5–6 October 2026). The companion accounts add “Recorder, Gauge Basin, Albany Filter Beds, (not quite completed)” for 1917 (Annual Report 1917, 428, https://hkintexts.histsyn.com/page/617399; accessed 5–6 October 2026) and the ledger line “Unfiltered water, Gauge Basin, Albany Road” for 1924 (Annual Report 1924, 440, https://hkintexts.histsyn.com/page/621442; accessed 5–6 October 2026).
  52. ↑Administrative Reports, Annual Report 1918, 440 (https://hkintexts.histsyn.com/page/617906; accessed 5–6 October 2026).
  53. ↑Colonial Office, CO129-252 – Acting Governor Barker & Governor Sir Robinson & Public Offices – 1891 [12], 240: “Watchman on the Pokfulam Conduit and our office becoming vacant. I do not propose to continue it but to arrange for the duties in connection with those of watchman at the Pokfulam Filter Beds.” https://hkintexts.histsyn.com/page/347224 (accessed 5–6 October 2026).
  54. ↑Hongkong Government Gazette, Public Works Return, January 1882, 1, head 3, “Maintenance of Water Works”: incidental services include “to repairing, covering stones Pokfulam Conduit, Water Mains, from old Glenealy Tank to Caine Road, Stop and Fire Cocks, Supply & Service Pipes, &c., to Government Buildings.” https://sunzi.lib.hku.hk/hkgro/view/g1882/678007.pdf (accessed 5 October 2026). https://hkintexts.histsyn.com/page/1665232 (accessed 5–6 October 2026).
  55. ↑Sessional Papers, Hong Kong, 1904, 288: a vote item covers the “Improvements- of Pokfulam Conduit” (OCR hyphen as indexed), and the report states that “the Pokfulam conduit has been thoroughly cleaned and freed from obstructions throughout its entire length.” https://hkintexts.histsyn.com/page/1824181 (accessed 5–6 October 2026).
  56. ↑Sessional Papers, Hong Kong, 1884, 26: “Vote of $950 for resumption by the Crown of a portion of Inland Lots 714 and 715 on Robinson Road, on account of the Pokfulam Conduit being under it.” https://hkintexts.histsyn.com/page/1812625 (accessed 5–6 October 2026).
  57. ↑Sessional Papers, Hong Kong, 1891, 143 (Public Works Department report, sewerage works): “considerable progress has been made … with the works of training the Nullahs in the Western District between the Pokfulam Conduit and the Robinson Road. The latter are nearly completed; the cost of these works to date has been $12,640.52.” https://hkintexts.histsyn.com/page/1815313 (accessed 5–6 October 2026).
  58. ↑Sessional Papers, Hong Kong, 1896, paras 52–53, 56–57: Chadwick’s report of 28 December 1889 divides the city into three zones, “the uppermost zone from the 250-feet contour to the Pokfoolum conduit”; the built area had “extended in places up to and even above the level of the Pokfoolum conduit”; the works were “finally decided upon and carried out during the years 1890-1892.” https://hkintexts.histsyn.com/page/1818321 (accessed 5–6 October 2026).
  59. ↑Antiquities and Monuments Board (Hong Kong), New Item Serial No. N132, N57, N257 & N251 (Pok Fu Lam Reservoir board paper), https://www.aab.gov.hk/filemanager/aab/en/content_4/20190319_historic_4_new_items.pdf (accessed 5 October 2026): reservoir construction from 1860, supply from the end of 1863, storage cut to 2,000,000 gallons, extensions 1866–1877; the paper also lists “The New Pokfoolum Waterworks”, The China Mail, 13 July 1878 among its sources.
  60. ↑Hongkong Government Gazette, 7 October 1899, “Legislative Council No. 12” (minutes of 4 October 1899), 3: proposed works to be referred to the Public Works Committee, item “(10) Pokfulam Conduit Road.” https://sunzi.lib.hku.hk/hkgro/view/g1899/630715.pdf (accessed 5 October 2026). https://hkintexts.histsyn.com/page/1637208 (accessed 5–6 October 2026).
  61. ↑Sessional Papers, Hong Kong, 1900, 244, para. (xviii): “Pokfulam Conduit Road.—Owing to several sites for residences being sold above the level of the Robinson Road sites, it became necessary to construct a road giving access to them. A commencement has been made with the approach road up Glenealy Ravine…” https://hkintexts.histsyn.com/page/1821098 (accessed 5–6 October 2026). The annual administration report for the same year adds that “Good progress was made on the Pokfulam Conduit Road during the year, and plans and estimates are ready for its completion as far as a junction with Robinson Road” (Hong Kong Annual Administration Reports, Colonial Reports–Annual, report for 1900, printed 302, https://hkintexts.histsyn.com/page/611000; accessed 5–6 October 2026).
  62. ↑Sessional Papers, Hong Kong, 1901, 338: “Pokfulam Conduit Road. This new road was carried as far as Inland Lot 1,549 during the year at a cost of $9,684.73.” https://hkintexts.histsyn.com/page/1821729 (accessed 5–6 October 2026).
  63. ↑Sessional Papers, Hong Kong, 1902, 953, para. 48: “Pokfulam Conduit Road.—This road was completed in November, with the exception of a short length near its western extremity … The width of the road as originally constructed was 12 feet … it was decided to increase it to 16 feet.” https://hkintexts.histsyn.com/page/1823279 (accessed 5–6 October 2026).
  64. ↑The Hongkong Directory & Chronicle for the Year 1904, 1391: residents’ addresses given as “Conduit Road” — the earliest Directory hit for the phrase in the corpus (a search of Directory_and_Chronicle_189* for Pokfulum returns zero). https://hkintexts.histsyn.com/page/1019819 (accessed 5–6 October 2026).
  65. ↑Executive Council minutes, CO131-(49,54), 1913–1920, 1088: acquisition of Inland Lot 1546 “known as No. 31, Conduit Road” and Inland Lot 1548 “known as Nos. 17, 19, 21, 23, 25, and 27, Conduit Road.” https://hkintexts.histsyn.com/page/5228917 (accessed 5–6 October 2026).
  66. ↑Corpus caveat. The Chinese-language press in this index begins about 1926 (daily papers about 1947), so the 1870s stratum is English-only; a silence in this paper about Chinese-language reporting is a silence of the corpus, not evidence of historical absence. Period place spellings are kept as printed (Pokfoolum, Pokfulum, Pokfulam, Pok Fu Lam). Modern Chinese forms were checked rather than reconstructed: Conduit Road = 干德道 and Albany Road = 雅賓利道 in the Lands Department gazetted street-name list; the Public Records Office = 香港政府檔案處 (grs.gov.hk).
  67. ↑Hongkong Government Gazette, 20 August 1914, no. 322, 318, “List of roads closed”: item 1, “Pokfulam Conduit west of Victoria Battery.” https://sunzi.lib.hku.hk/hkgro/view/g1914/35981.pdf (accessed 5 October 2026). https://hkintexts.histsyn.com/page/1603857 (accessed 5–6 October 2026).
  68. ↑Colonial Office, CO129-335 – Governor Nathan – 1906 [8-10], 292: “Landslips threatened but did not seriously injure the Pokfulam Conduit.” https://hkintexts.histsyn.com/page/405750 (accessed 5–6 October 2026).
  69. ↑Administrative Reports, Hong Kong, Annual Report 1922, 610: “A new 2″ main 4,500 lin. feet in length was laid from Pokfulam Conduit to supply Houses on the New Mount Davis Road … A new 2″ main 3,000 lin. feet in length was laid from Pokfulam Conduit to houses in the ‘Villa Miramere’ District.” https://hkintexts.histsyn.com/page/620195 (accessed 5–6 October 2026).
  70. ↑Colonial Office, CO129-598-2 Tai Lam Chung Reservoir Project 24-6-1948 - 3-3-1949, 43: the colony’s trunk mains are listed as “(a) The Pokfulam Conduit, (b) The Tytam tunnel and Bowen Road Conduit, (c) The Aberdeen main, (d) The Shing Mun-Kowloon-Hong Kong mains”, followed by “There is no doubt that the brick conduits have given excellent service, but their capacity is limited to the free discharge on very flat gradients, for being covered only with loose stone slabs they do not permit of any pressure head.” https://hkintexts.histsyn.com/page/584000 (accessed 5–6 October 2026). The same passage says the four trunk mains were damaged “one after the other” in the 1941 attack without naming which, when or how badly; no damage to this conduit is established.
  71. ↑CO129-598-2, 43 (https://hkintexts.histsyn.com/page/584000; accessed 5–6 October 2026).
  72. ↑Antiquities and Monuments Board, N372 Pok Fu Lam Conduit (accessed 5 October 2026).
  73. ↑On negative results: no corpus document names the drawing, and the 1891 Gazette notification is unattested (doc search as above).
  74. ↑Government Records Service (Hong Kong), catalogue record HKRS1951-1-3 (file dates 1871–1877; accessed 5 October 2026).
  75. ↑CO129-168, 220 (https://hkintexts.histsyn.com/page/294250; accessed 5–6 October 2026).
  76. ↑China Mail, 13 July 1878, 6 (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026).
  77. ↑China Mail, 13 July 1878, 6 (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026).
  78. ↑Gwulo, “The Pokfulam Conduit” (accessed 5 October 2026).
  79. ↑CO129-193, 163 (https://hkintexts.histsyn.com/page/308823; accessed 5–6 October 2026).
  80. ↑On the completion date: Sessional Paper 1896 (“completed in 1877”) against China Mail, 13 July 1878 (as above).
  81. ↑On negative results: no corpus document names the drawing, and the 1891 Gazette notification is unattested (doc search as above).
  82. ↑Antiquities and Monuments Board, N372 Pok Fu Lam Conduit (accessed 5 October 2026).
  83. ↑Corrections recorded during verification. First, a remembered citation to a 1910 Public Works Department report (https://hkintexts.histsyn.com/page/613600; accessed 5–6 October 2026) as describing a gauge basin was re-fetched on 5 October 2026 and found to be a rainfall paragraph (Annual Report 1910, 147); it was dropped. Second, four search labels (pf1875, pf1876, pf1877, pf1878) were issued without a document_key filter, so all four returned the same unfiltered total and overwrote the same result file; no year-sliced count is quoted in this paper, and any such count must be re-run with a regexp or prefix filter first (Appendix A).
  84. ↑Corrections recorded during verification, as in the full note above: the retraction of the misremembered 1910 citation (https://hkintexts.histsyn.com/page/613600; accessed 5–6 October 2026) and the four invalid year-sliced searches.
  85. ↑CO129-598-2, 43 (https://hkintexts.histsyn.com/page/584000; accessed 5–6 October 2026).
  86. ↑Negative control: CO129-171, 300 (https://hkintexts.histsyn.com/page/296570; accessed 5–6 October 2026), describes the Low Level (Tytam) Project — “Before entering the latter, the water will be passed through a small quage basin furnished with a proper guage” and “After passing through the Guage the water enters the Conduit, which will be 2 feet wide by 2 feet deep in cross section. The Conduit is 1356 yards long…” These are Tytam works and have nothing to do with the sheet under study; the passage is cited only so that it is not mistaken for a description of the Pokfulam gauge basin.
  87. ↑Price, “Report on the Water Supply of Hongkong,” Gazette, 10 July 1875, 283 (https://hkintexts.histsyn.com/page/1673080; accessed 5–6 October 2026) — the High Level and Low Level Tytam estimate tables.
  88. ↑Price, “Report on the Water Supply of Hongkong,” Gazette, 10 July 1875, 282 (https://hkintexts.histsyn.com/page/1673079; accessed 5–6 October 2026).
  89. ↑Antiquities and Monuments Board, N372 Pok Fu Lam Conduit (accessed 5 October 2026).
  90. ↑On what is claimed: subject, location and old/new relation only; the date, authorship and “twenty sheets” question remain unestablished (full note above).
  91. ↑On negative results: no corpus document names the drawing, and the 1891 Gazette notification is unattested (doc search as above).
  92. ↑Corpus caveat. The Chinese-language press in this index begins about 1926 (daily papers about 1947), so the 1870s stratum is English-only; a silence in this paper about Chinese-language reporting is a silence of the corpus, not evidence of historical absence. Period place spellings are kept as printed (Pokfoolum, Pokfulum, Pokfulam, Pok Fu Lam). Modern Chinese forms were checked rather than reconstructed: Conduit Road = 干德道 and Albany Road = 雅賓利道 in the Lands Department gazetted street-name list; the Public Records Office = 香港政府檔案處 (grs.gov.hk).
  93. ↑China Mail, 13 July 1878, 6 (https://hkintexts.histsyn.com/page/1890648; accessed 5–6 October 2026).

Bibliography

Artefact and catalogue

  • Government Records Service (Hong Kong). Catalogue record HKRS1951-1-3, Hong-kong Waterworks – Pokfoolum Conduit. New Gauge Basin, series HKRS1951 Drawings Relating to Water Works, original ref W00239, file dates 1871–1877. https://search.grs.gov.hk/en/arcview.xhtml?q=HKRS1951-1-3&eid=hlz6QKooYl3kDojwfO8Erw%3D%3D&ls=e_k%3D1%26q%3DHKRS1951-1-3%26rpp%3D10.
  • [Artefact] HONG-KONG WATERWORKS. POKFOOLUM CONDUIT. NEW GAUGE BASIN. Ink and wash on paper, 20 feet = 1 inch, sheet mark “W 239.” Public Records Office (Hong Kong), HKRS1951-1-3; digital file Pokfoolum Conduit/HKRS1951-1-3_0001.jpg (3043 × 2048 px). Read 5 October 2026.

Government Gazette

Colonial Office records (CO129, CO131)

Sessional Papers, Administrative and Public Works reports

Newspapers and directories

Modern sources (corroborative)

  • Antiquities and Monuments Board (Hong Kong). N372 Pok Fu Lam Conduit Public Consultation. https://www.aab.gov.hk/filemanager/aab/en/content_4/N372_Public%20Consultation%20(ENG).pdf.
  • Antiquities and Monuments Board (Hong Kong). New Item Serial No. N132, N57, N257 & N251 (Pok Fu Lam Reservoir). https://www.aab.gov.hk/filemanager/aab/en/content_4/20190319_historic_4_new_items.pdf.
  • Gwulo (Old Hong Kong). “The Pokfulam Conduit,” 11 March 2011. https://gwulo.com/pokfulam-conduit.
  • Industrial History Hong Kong. “Hong Kong Water Supply – Pok Fu Lam Reservoir.” https://industrialhistoryhk.org/hong-kong-water-supply-pok-fu-lam-reservoir/.
  • Lands Department (Hong Kong). Street name list — Conduit Road 干德道, Albany Road 雅賓利道.
at