CHAPTER VI. New River Head
The site south-west of Sadler's Wells known as New River
Head takes its name from the reservoir at the mouth of the
New River, the channel cut in 1604–13 to supply London
with water from springs in Hertfordshire—a civil
engineering achievement vital to the development of the
metropolis. From here, high in the fields of rural
Clerkenwell, a network of wooden mains conveyed water
to the cisterns of London. To begin with New River Head
consisted merely of a reservoir, the Round Pond, and a
single building known as the Water House. In time it
became a comparatively sophisticated complex with outer
ponds and associated structures covering seven acres—a
roughly trapezoidal site, bounded latterly by Rosebery
Avenue and Amwell Street on the east and west,
Myddelton Passage and Hardwick Street on the north
and south.
This has long ceased to be the place of watery tranquillity evoked by early views (Ills 208, 209, 215), and,
with its disparate buildings grouped around a park-like
open space, it is no longer even obviously the site of a
waterworks (Ills 211, 212). Nor is its present largely residential character immediately apparent. The principal
buildings—neither fronting Rosebery Avenue—are the
former Metropolitan Water Board offices of 1915–20,
which incorporate a fine late seventeenth-century room
from the Water House, and the board's water-testing laboratory of 1936–8. Both were converted into flats in the
1990s. With the construction in 2000–3 of two further
blocks of flats, and the landscaping of much of the open
space as gardens, including a grass deck over a car-park on
what was once a filter bed, the greater part of the site
became private and residential.

208. London from Islington Hill, by Thomas Bowles, c. 1740. New River Head, centre-left, Upper Pond in foreground
A sense of the corporate and civic institutional history
of the place is conveyed by the architectural character of
the older buildings, by the fact of public access to parts of
the gardens and, more explicitly, by a public viewing area
with interpretative panels, which marks the start of a footpath along the course of the New River itself. Substantial
nineteenth-century revetments evoke something of the
scale of the bodies of water once contained here, and there
are two remarkable survivals connected with the pumping
of water to an outlying reservoir at Claremont Square: the
stump of an early eighteenth-century windmill and an
imposing engine house, built and enlarged in the late
eighteenth century. New River Head's historic role in the
supply of water to the capital is perpetuated, unobtrusively, by a small pumping station and deep access shaft,
components in the London Ring Main completed in 1994.

209. New River Head from north, c. 1910. Water House, right of centre, at margin of Round Pond, filter beds in foreground
The New River
The complex history of the New River has been well
described elsewhere and will not be repeated here, but
some discussion of the formation of the 'river'—in fact an
aqueduct or conduit in canal form—is required to set the
scene for an account of the development of New River
Head. Shortages in the supply of clean water for London
became a persistent concern in the sixteenth century, as
the city grew and existing rivers, streams, wells and conduits became inadequate and polluted. Since the thirteenth century conduits had carried water to London
from springs in rural districts to the north, including
Clerkenwell. More such conduits, from Hackney,
Hampstead and Muswell Hill, were contructed by the City
Corporation after an Act of 1543. In 1582 the London
Bridge Waterworks began to pump London's first regular
water supply direct to private houses, but the operation of
its water wheel, under the northernmost arches of the
bridge where the tidal Thames ebbed and flowed, was
irregular, and the water impure. (ref. 1)
Other proposals for improving the City water supply
were considered throughout the 1590s, but without immediate effect. Decisive action towards the realization of a
project on a grand scale came in 1604, the year after a
plague epidemic killed about 30,000 Londoners. James I
granted Captain Edmund Colthurst, a former army officer
from Bath, a charter allowing him to bring 'sweet' spring
water from Hertfordshire to 'particuler howses and places'
in London in a specially cut channel, plans for which
Colthurst had been promoting since at least 1602. By 1605
Colthurst had evidently cut two or three miles of waterway at the Hertfordshire end of what was to become the
New River. (ref. 2) The City of London then belatedly engaged
with the implications of Colthurst's project, and gained
an Act of Parliament authorizing the Corporation or its
deputies to bring water to London from the Hertfordshire
springs at Chadwell and Amwell near Ware, allowing compensation to Colthurst for any resultant losses.
One of the MPs on the Parliamentary committee that
considered this and alternative schemes was Hugh
Myddelton, an eminent London goldsmith, merchant and
entrepreneur. Myddelton's nascent interest in the Act's
outcome may have been represented by William Inglebert,
who proposed placing the intended waterway in an
enclosed brick conduit. That expensive possibility was
sanctioned by a second Act in 1606 but never seriously
pursued. Complex negotiations ensued, with Colthurst at
the centre. He needed funds to continue, but the City was
unwilling to make a financial commitment. The upshot of
what remains a murky story of the confluence of private
enterprise and public policy was that Colthurst was
replaced as the scheme's promoter in March 1609,
Myddelton becoming the City's deputy in the implementation of the powers granted in 1605. Myddelton and his
partners took on the costs of creating the waterway and
secured the eventual profits of facilities that they were to
build, own and operate. Colthurst's displacement was
probably by mutual agreement, as he remained involved as
overseer of the works, and received a substantial shareholding on favourable terms. (ref. 3)
Construction of what was henceforward known as the
New River recommenced in late 1609. Working from the
north, a channel 10 ft wide and 4 ft deep was dug along an
ingeniously devised course that maintained a gradual and
constant fall of five inches in the mile. Nearly forty miles
of meandering canal were needed to cover the distance of
just over twenty miles as the crow flies. Within months
opposition from landowners along the course had stopped
progress. After a protracted and costly dispute, in 1611
James I rescued Myddelton and his partners, agreeing
to take on half the costs of the project in return for half
the profits. (ref. 4) A private speculation with ambiguous state
support thus became a firm partnership, effectively underwritten by the Crown—a significant step towards 'a new
acceptance of the role of commerce in the running of the
civic state'. (ref. 5) Work proceeded at an intense rate from
January 1612 with Edward Pond, a mathematician and
almanac maker, as surveyor. (ref. 6) The Clerkenwell terminus at
'Islington Hill' was built in the summer of 1613, and
Myddelton staged a ceremonial opening at New River
Head on Michaelmas Day 1613: 'the flood-gates flew open,
the streame ranne gallantly into the cisterne, drummes
and trumpets sounding in a triumphall manner'. (ref. 7) Soon
water was flowing through wooden pipes to households in
the northern and western parts of the City. (ref. 8)
Nothing of the New River itself survives in
Clerkenwell. It flowed into the Round Pond at New River
Head from the north-east, the final stretch running along
what is now Rosebery Avenue. For a long time it passed in
front of Sadler's Wells, contributing to the latter's appeal
as a resort (Ills 210, 211, and 187 on page 149). From an
early date the New River's Hertfordshire springs were
supplemented by the River Lea, and by the 1850s New
River water was almost entirely Lea water. (ref. 9) In 1946 the
New River, by this time running underground, was truncated to terminate at Stoke Newington.
The New River Company and its successors
In 1619 the status of the proprietors of the New River was
secured through incorporation by royal charter as 'The
Governor and Company of the New River brought from
Chadwell and Amwell to London'; Myddelton, who was
made a baronet in 1622, was the first Governor. Profits did
not come quickly and in 1631 Charles I sold the Crown's
half of the undertaking for an annual payment of £500, a
poor bargain for the King as within a few years his annual
dividend would have exceeded this sum. (ref. 10)
The company's senior officer was the Clerk, who contracted to maintain the river for a fixed fee. From 1667 to
1705 this office was held by John Grene, a wealthy shareholder, who took up residence in the Water House. Initially
peripatetic in its gatherings, the company found settled
premises at Puddle Dock in the City by 1669, moving in
1717 to a house in Bridewell Precinct, with a wharf where
elm logs for water pipes were landed. These offices, together
with many early records, were destroyed by fire in 1769.
New headquarters were built in 1770–1 on an adjacent site
at Dorset Garden. With the introduction of iron watermains after 1810 the timber wharf became redundant, and
in 1820, having survived a period of intense competition in
good shape, the company moved its headquarters to the
newly enlarged Water House at New River Head. (ref. 11)
London's water supply came under increasingly close
official scrutiny from the 1840s, and pressure for its
municipalization began to build soon after the Second
Reform Act, in 1867. (ref. 12) This led ultimately to the appointment of a Royal Commission in 1897 and to the formation of the Metropolitan Water Board in 1902 to take over
London's private water companies, the handovers taking
place in 1904. The significance of New River Head in the
history of London's water supply was reinforced when, in
1913, the board decided to locate its headquarters there.
The bulk of the New River Company's Clerkenwell estate,
no longer part of the water undertaking, remained in
private hands through the New River Company Limited,
made up of former shareholders. (ref. 13)
In 1973–4 the Metropolitan Water Board was abolished
as part of a national reorganization of the water industry.
A new authority, Thames Water, took responsibility for
water supply in London and a much wider region, moving
its headquarters from New River Head to Reading in 1987.
Privatization followed in 1989, and in 2001 Thames Water
plc became a subsidiary of the German energy conglomerate the RWE Group. Thames Water was sold again in
October 2006 to a group led by Macquarie Bank of
Australia. (ref. 14)

210. New River Head, 1743

211. New River Head, c. 1874
Development of New River Head since 1613
Northern Clerkenwell was chosen for the termination of
the New River for a combination of geographic and pragmatic reasons. The flow of water from the reservoir to the
City was to depend only on gravity, so the reservoir needed
to be on high ground yet close to the metropolis. Further,
while much of southern Clerkenwell is free-draining
gravel, the New River Head site is underlain by more
impervious London clay, providing a better bed for a
reservoir. The spot chosen was as elevated as the gradient
from Hertfordshire would allow, and already had a pond,
hitherto used for setting dogs on to ducks. The freehold
of the land, generally let for grazing, was owned by Sir
Samuel Backhouse, one of the New River's original
'Adventurers' or shareholders. Backhouse was amply compensated, and his involvement may be as important as any
other factor in explaining why this particular site was
chosen. (ref. 15)
In anticipation of the arrival of the water in 1613 the
Round Pond was formed, 200 ft in diameter and lined with
oak campshedding. It was encircled by a brick security
wall, built by Stephen Boone, bricklayer, one of the New
River's principal builders. Once the pond had been filled,
outward flow towards the City was controlled through a
cistern and stopcocks in the basement of the Water House
on the pond's southern edge. (ref. 16) From the beginning, waste
or overflow water was discharged from the end of the New
River outside the Round Pond. It was intended that this
would drain via a ditch to the south-west into the River
Fleet, but the overflow soon generated its own 'waste
pond', extending amorphously over more than an acre
north-east and south-west of the Round Pond (Ills 210,
215). (ref. 17) The Outer Pond, as it became known, continued to
grow until it was given more definite shape by excavations
over an area of more than two acres. By 1730 the margins
had been embanked and an outer cistern-house built; an
inscription reading 'wharf built 1726, pond cleaned' on a
then loose stone was recorded in 1948. (ref. 18) There was also a
'middle' cistern house south-west of the Water House (Ills
208, 210). All the while the Outer Pond was one of the
'principal places for angling near London'. (ref. 19)
By the 1670s the supply of water to the West End had
come within the New River Company's purview. The first
decade of the eighteenth century saw initiatives to extend
and reinforce the company's reach, against a background
of competitive pressures and falling dividends. In 1708 an
additional high-level reservoir, the Upper Pond, was built
away to the north at what is now Claremont Square—the
highest ground locally, immediately to the east of the site
of the Civil War 'Fort Royal' of 1642–3. This gave a better
head of pressure, allowing more distant and higher-lying
areas in and around the West End to be supplied. (ref. 20) The
power required to pump water to the Upper Pond from
the Round Pond was supplied at first by a windmill, built
beside the Outer Pond to the north-west, but this did not
work well, and was eventually succeeded by an atmospheric steam engine, at work from 1768.

212. New River Head, 2005
Most of the land around New River Head remained
open fields for longer than adjoining landholdings, significant parts being given over to the accretive network of
pipes radiating from the Round Pond. The New River
Company's status as a major Clerkenwell landowner was
cemented only in 1744 when it acquired 44 acres around
New River Head and the Upper Pond. The Round Pond
apart, New River Head remained an essentially open site
enclosed only by low timber fences. These fences began to
be replaced in 1770, largely by high brick perimeter walls,
secure enclosure of the compound being completed in
1780 following the Gordon Riots, during which troops
were stationed at the waterworks. (ref. 21) As the demand for
New River water grew, several other small reservoirs were
formed in the locality. A West Pond was built in 1779–81,
another by 1805, to serve as nodes between mains for the
supply of northern suburbs. One of these was where
Archery Fields House now stands, near Lloyd Square, the
other slightly to the north on the line of Great Percy
Street. Supply to eastern parts of the City was improved
by the construction of the New or St John Street
Reservoir in 1805, an oval pond lying between Sadler's
Wells and the west side of St John Street (see Ill. 99 on
page 87), where there was an associated circular cistern
house. An oblong pond where Joseph Trotter Close now
stands was probably dug about 1760 in connection with
the English Grotto (see page 89). It was taken back by the
New River Company in 1781 as a reservoir, and filled in
c. 1810 prior to the development of Myddelton Street. (ref. 22)
The New River Company's first 'engineer and surveyor'
was Myddelton's kinsman Henry Mill, who held the post
from c. 1718 and lived in the Water House. He was probably responsible for the reconstruction of the Outer Pond.
Incapacitated by a stroke, the elderly Mill was assisted
from 1767 by Robert Mylne, whose appointment as joint
surveyor probably arose through his responsibility for
Blackfriars Bridge, which adjoined the company's wharf.
Mylne designed the new Dorset Garden headquarters and
became sole surveyor in 1771, retaining the post until his
own dotage in 1810. He too lived at the Water House,
dying there in 1811. (ref. 23)
Intensifying competition in the supply of water across
the rapidly growing metropolis brought about important
developments in the early nineteenth century, as the New
River Company sought to safeguard its position. The
replacement of wooden pipes by cast-iron mains, complete
by 1819, allowed building on the company's land (see
Chapter VII). These initiatives were overseen by William
Chadwell Mylne, whose middle name speaks of his
father's commitment to the New River. He shared the
company's surveyorship with his father from 1804, taking
over the post (together with the tenancy of the Water
House) in 1810, and retaining it until 1861. (ref. 24)
In 1842 W. C. Mylne relined the Round Pond, replacing the decayed campshedding with fender piles and castiron plate wharfing, in line with recent innovations in the
use of iron for harbour and riverside quay walls. The supplier was a Mr Ward. (ref. 25) Most of the iron wharfing around
the northern bank of the pond survives, many of the castiron plates retaining moulded 'cornice' bars (Ills 213, 214).
Growing demand for water, poor supply and, above all,
rising concerns about impurity, led to the Metropolis
Water Acts of 1852 and 1871. These required filtration
through sand, covered reservoirs and the provision of a
constant or on-demand supply rather than what had
hitherto been an intermittent supply, controlled by
water-company turncocks. The 1852 Act led directly to
reconstruction at New River Head, with the Outer Pond
re-formed as filter beds, and to the covering of the
Claremont Square reservoir (see pages 193–5). These
works, required to be done within five years, were carried
out to Mylne's plans in 1854–6. At the same time, the
Round Pond was given a sloping revetment or retaining
wall within its iron sides, to give greater support to the
banks. The upper face, which sloped down to an inner
dwarf brick wall, and the bottom of the pond were paved
with York-stone slabs, to prevent the water becoming too
muddy. All the New River Head building work was carried
out by George Mansfield & Son. (ref. 26)
From the newly reinforced Round Pond water passed
to the three new filter beds along radiating channels before
percolating down through ever-finer layers of gravel and
sand to holes in the brick-paved floors. The Round Pond
could thus remain open to the elements, as it was not
deemed to be a reservoir, but rather a collecting pool for
feeding the filter beds (Ill. 209). (ref. 27) Like the cast-iron wharfing, the Round Pond revetment survives around the northern arc. The margins of the western filter bed are also still
recognizable, the site remaining sunken with car-parking
under a private communal garden known as 'The Garden
Deck'. Around the south and west perimeter the Yorkstone-faced and bullnose-capped revetment of the 1850s
can still be seen.

213. Cast-iron lining of Round Pond, c. 1914–15, with section
through slope of later revetment. In left foreground an early
wooden water pipe can also be seen

214. Looking west across New River Head in 2005. In foreground and centre surviving portion of retaining structure of
Round Pond; centre-left, Devil's Conduit; behind, engine
house; right, the Nautilus Building (Nicholson GDA,
architects, 2001–3)
Clearance of the Water House and drainage of the
essentially redundant Round Pond in 1914–15 created
space for the Metropolitan Water Board's central offices.
Since the formation of Rosebery Avenue in the early
1890s the site had once again been open to public view,
and railings were erected in the 1920s in lieu of the lost
perimeter walls. Similar replacement followed along
Amwell Street in 1933. In 1936–8 the eastern filter bed
was built over for a water-testing laboratory and a garden
with a fountain pond. Associated new points of access
were created on Rosebery Avenue and Arlington Way,
with gates and flanking sections of somewhat more ornamental railings. The rest of the mid-nineteenth-century
system continued in use until 1946. Thereafter the central
and western filter-bed sites were laid to lawn and then
used for car-parking and prefabricated temporary
offices. (ref. 28)
The north-west corner of the New River Head site is
occupied by Charles Allen House, a seven-storey block of
fourteen three-bedroom flats facing Amwell Street. This
was built in 1964–6 to designs by J. F. Hearsum, Surveyor
to the Metropolitan Water Board, to house employees previously accommodated in a miscellany of cottages on this
side of the site. (ref. 29)
To the south, also facing Amwell Street, is a pumping
station for the London Ring Main constructed in
1986–94. A 50-mile concrete tunnel carrying some 250
million gallons of water around the metropolis, the Ring
Main brought new purpose, or return of old purpose, to
New River Head. A small stock-brick control or
switchgear building provides below-ground access to the
head of one of the seventeen access shafts, where six powerful Weir motors, ranging from 315 kW to 626 kW capacity, each pump up to 13.2 million gallons a day. The
modest outward appearance of these structures belies
their significance.
Following the relocation of Thames Water's headquarters to Reading, and privatization in 1989, New River
Head became largely redundant. The nature of the place,
with its large open spaces and mixture of historic buildings, ruled out complete redevelopment. A planning brief
was drawn up in 1991, and Thames Water set about realising the value of the site. Once the former central offices
and water-testing laboratory had been made residential,
two five-storey blocks of flats were built on peripheral
parts of the site for St James Homes, a joint venture by
Thames Water and Berkeley Homes. The architects were
Nicholson GDA. To the north is the Nautilus Building of
2000–1 at No. 3 Myddelton Passage, with fifty-two flats
(Ill. 214). This echoes the design of the former laboratory
adjoining, reversing the curves of the earlier façades. To
the south-west the Hydra Building of 2001–3 at No. 10
Hardwick Street has thirty-six flats, of which twenty were
'affordable' or social housing, in conformity with Greater
London Authority stipulations. The two-storey stockbrick building immediately to its east, at No. 9 Hardwick
Street, is the former water-meter testing house of 1922–4.
Renamed the Remus Building, this provides eight more
flats. The space to the south of the Nautilus Building has
been landscaped as a formal garden with a fountain, open
to the public and entered via the former offices, or through
a gate on Myddelton Passage. An adjoining gate gives
access to the small viewing area at the start of the New
River walk. (ref. 30)

215. New River Head from north in 1665. Enclosure of Round
Pond and Water House of 1613, beyond Outer or Waste
Pond. Engraving by Wenceslaus Hollar

216. New River Head from the south, 1730–1, by Bernard Lens. Water House (1613) to right, disused windmill to left
The Water House
The first building at New River Head, the Water House,
stood for three hundred years, latterly in much extended
form. The original brick 'cestern house' of 1613 was built
by Stephen Boone, likely to have been an individual of
standing, given the project's profile; its architect, and the
building's appearance, remain unknown. It punctuated the
Round Pond's security wall and had a dual purpose. Above
its basement 'gallery', from which stopcocks regulated the
flow of water from the pond into pipes, there was a tall
lodge, rising two full storeys to a high, near-pyramidal roof
(Ills 215, 216). In an almost square plan the main block
had a single room on each level, with a ground-floor
counting house, a first-floor 'middle room', and garrets
providing accommodation for the site supervisor, initially
Howell Jones. On the outer or south front there were fullheight (giant order) engaged columns, under a deep classical eaves cornice, in an elevation that is most readily
associated with the 1630s. No rebuilding is recorded, so
this does appear to be an exceptionally early example of
such treatment. On the pond side there was a square projecting stair turret through which the building was
entered. (ref. 31) Jones appears to have used the building for a
third and unofficial purpose; after his death the New River
Company resolved in 1622 'that none shalbe there suffered
hereafter to sell drinck or kakes, or to use victualling in
any sort whatsoever'. (ref. 32) Subsequent resident supervisors
included Edward Sadler, probably the founder of Sadler's
Wells, who, in 1664 or 1665, was permitted to continue to
live at the Water House. By 1675 William Markeham,
turncock, was in the building, paying £5 annual rent, and
Sadler was living near by as a victualler. By 1691–2 James
Mathers was in the Water House. (ref. 33)

217. Water House from north in 1824. As enlarged by Robert
Mylne (1778–89). Wings to left added by William
Chadwell Mylne in 1818–20
The status of the Water House changed in 1693 when
it was enlarged by and for Hugh Myddelton's grandsonin-law John Grene, the company's wealthy Clerk. (ref. 34) He
added rooms on the first floor supported on Doric loggias,
to either side of the stair turret, which appears to have
been newly fenestrated (Ill. 218). These additions were
secondary to a sumptuous remodelling of the 'middle
room', with carved panelling and a decorative plaster
ceiling, to form what subsequently came to be known as
the Oak Room (Ills 229, 231–2). The rooms over the
loggias were also given decorative ceilings. With windows
on three sides, the new Oak Room allowed views over the
fields to London, and was evidently open to curious sightseers. Writing of New River Head and its water-pipes
('canaux') in his guide to London of 1693, Colsoni added
that 'si vous voulez vous pouvez entrer dans cette Maison
là & y voir les dits canaux'. (ref. 35) The Oak Room, one of the
loggia ceilings and a few other features survive today in
the former Metropolitan Water Board offices on the site
of the Water House (see below).

218. New River Head from the north, 1730–1, by Bernard Lens, showing Water House as enlarged in 1693
In the eighteenth century the Water House became the
residence of the New River Company's Surveyor, first
Henry Mill, then, from 1771, Robert Mylne, who
repaired, refaced and enlarged the building in, respectively, 1778, 1782 and 1789. This work involved infilling
the loggias as well as the addition of upper-storey rooms
and the removal of the pyramidal roof, ingeniously generating a more conventional tripartite façade to the pond.
The seventeenth-century building, while essentially
intact, had been outwardly altered beyond recognition; the
lodge had taken on the form of a suburban villa (Ill. 217). (ref. 36)
Two inscribed stones from Mylne's refacing, recording the
dates 1613 and 1782, have been reset in tympana over
adjoining first-floor windows to the relocated Oak Room
on the west side of the former Metropolitan Water Board
offices. W. C. Mylne extended the Water House eastwards
in 1818–20, for the relocation of the company's 'Court
House', or administrative offices. (ref. 37) Following the younger
Mylne's retirement in 1861 the building was given over
entirely to office use. In 1893 a further large five-storey
southern extension, facing the newly made Rosebery
Avenue, was built by Dove Brothers (Ill. 209). (ref. 38) What had
become a rambling complex was demolished in 1914–15.

219. New River Head engine house, as built in 1766–8 (John
Smeaton, engineer), from the north-west c. 1780. Behind, the
former windmill of 1707–8
The Windmill and Engine House
A low, round, red-brick shed with a conical roof is all that
now remains of the windmill built in the early years of the
eighteenth century to pump water to the Upper Pond, and
so provide the head of water needed to supply Soho and
other high-lying parts of the West End. The mill, a notable
engineering experiment, was the first step in what was to
be a long search for efficient pumping machinery.
In the 1690s financial difficulties and intensifying competition began to threaten the New River Company's profitability, and the situation worsened after 1701–3, when
the London Bridge Waterworks acquired new pumping
equipment designed by George Sorocold, the first great
English water-supply engineer. Equivalent improvements
were deemed a necessary response, even though in 1702
Sir Christopher Wren advised the company against the
use of a pumping engine. Two years later the Rev. John
Lowthorp, another fellow of the Royal Society, set out the
alternatives of wind, horse or water power for pumping,
advising against a windmill and preferring a waterwheel. (ref. 39)
Lowthorp and others who submitted ideas found no
favour. The company instead turned to the highly reputed
Sorocold. In 1707 it adopted his experimental and expensive scheme for a windmill and an upper reservoir, on land
belonging to the 2nd Earl of Clarendon, a shareholder,
who had inherited the Backhouse land through marriage
in 1666. This would not only improve supply to the West
End, but also open up the possibility of supplying water
to Islington. The windmill, built in 1707–8 to the northeast of the Outer Pond, powered four pumps to force
water from the Round Pond to the new Upper Pond, a rise
of about 30 ft over a distance of about 1,000 ft. It consisted
of a tapering round brick tower, then rare in England, with
oval windows, an ogee cap, and six rather than the more
common four sails (Ills 208, 216). At the base was an integral horse gin for use when there was little wind.

220. New River Head engine house, section as designed by
Smeaton in 1766–7. From left to right, the boiler, cylinder,
beam and pump

221. New River Head engine house, ground-level plans in 1768, 1786, 1795 and 1849. Original engine house shaded

222. New River Head engine house, west elevation in 1768 and as enlarged by 1849. Top right and bottom left, details of south
boiler-house roof truss and staircase of 1848–9
New River Head engine house
The high-level pond was a success, but the windmill
was not. Possibly too sheltered, it did not work well and
was abandoned in 1720, the sails being removed, perhaps
following storm damage. Even more ignominiously, the
base, at just under 23 ft in diameter, proved too small to
serve well for the horse gin. Consequently a new building,
later known as the 'square horse-works', was erected
c. 1720, immediately adjoining the defunct windmill to the
south-east; for nearly fifty years this alone supplied the
Upper Pond. (ref. 40) Around 1770 the mill tower was cut down
to about two storeys and castellated, being further reduced
to its present form by the mid-nineteenth century, to serve
as a workyard store (Ill. 223). (ref. 41)
Elsewhere in London steam power was beginning to be
applied to water supply even as Sorocold's sails were
turning. The York Buildings Waterworks tried a Savery
'fire engine' c. 1710, replacing it with the first Newcomen
atmospheric steam engine to be used for water supply in
1726; neither was a success. The Chelsea Waterworks,
established in 1723, equipped itself with two Newcomen
engines in 1742 and 1747 that did work effectively. (ref. 42) By
comparison with these competitors the New River
Company came to steam relatively late. What is extraordinary is that its first steam-engine house, built in 1766–8
and certainly an early example of this building type, still
stands, albeit enveloped by later additions (Ill. 221). With
demand for New River water in the West End continually
increasing, John Smeaton, who had established himself as
a leading engineer in the 1750s, was commissioned in 1766
to investigate methods of improving supply to the Upper
Pond. He recommended a steam engine that would more
than double the pumping power of the horse works and,
he suggested, be cheaper. Designs for a Newcomen-type
beam engine were prepared in 1767, incorporating experimental improvements. The steam engine, Smeaton's first,
was erected in 1768 to the north-east of the former windmill, under the supervision of the recently appointed
Robert Mylne. At work by early 1769, it had an asymmetrically pivoting beam, of laminated timber, enabling an
unusually long stroke in a 9 ft-long cylinder of 18 in.
diameter (Ill. 220). (ref. 43)
Smeaton's engine house (Ills 219, 222) was a tall, heavily
buttressed slab, its short side 'greatly resembling a church
tower', (ref. 44) more so as it lacked the intended pediment. The
cylinder was housed in the larger southern chamber, and
the pump in the northern chamber. Two round boilers
were installed in a lean-to on the south side. The south
wall of the engine house of the 1760s is still outwardly
visible at the centre of the enlarged building, the original
east, west and internal or 'bob' walls largely surviving
within. (ref. 45)
By 1774 it had become clear that Smeaton's engine did
not work anything like as efficiently as had been hoped.
Like Sorocold's, Smeaton's record at New River Head is
one of trial and failure. However, the setback prompted
him to experiment further with steam engines, work that
significantly informed the innovations of James Watt. In
the meantime the New River Company and Mylne had to
think again, deciding to improve supply to the Upper
Pond in 1776. By 1779 an overshot water wheel had been
built, wholly underground on the square horse-works site
south-east of the former windmill. This supplementary
power source, which had a run-off to the West Pond of
1779–81, was used until c. 1850. (ref. 46)

223. New River Head engine house, from the west, c. 1910.
Stump of windmill to right
In 1782 the New River Company approached Boulton
& Watt with a view to commissioning a new engine to
stand alongside Smeaton's. Matthew Boulton and James
Watt had been manufacturing vastly more efficient steam
engines since 1775, and in 1782 Watt patented doubleacting and rotative engines. Designs for a machine incorporating these latest refinements were prepared, but this
was not built, though a similar engine was erected at
Chelsea Waterworks. Agreement for the New River Head
engine was not finally reached until February 1785, when
Boulton & Watt undertook to supply a double-acting
engine with a 32 in. cylinder and 8 ft stroke, working by
parallel rather than rotative motion. This was at work by
the end of 1786, in Mylne's westward extension of the
engine-house. Buttresses were again included, to reduce
vibration, and survive on the building's west side. The
new pump chamber, again to the north, also housed a
staircase and a high-level cistern. The Boulton & Watt
engine was about three times as efficient as Smeaton's,
which had been deemed wholly useless by 1792. (ref. 47)
Redundancy, technological advance and ever-increasing
demand provided scope for further improvement. In 1793
Robert Mylne gained approval for the acquisition of a
second engine which, working only occasionally, would
double the quantity of water raised. There was no
urgency, Mylne explaining to Boulton that 'I have begun
this business rather early, on purpose to have full time for
thought, afterthoughts and Pentimento's'. (ref. 48) Accordingly,
work on this improvement appears to have continued into
1797. It involved replacement of the Smeaton engine and
further extension of the engine house, this time on the east
side, for the installation of a double-acting parallel-motion
engine with a 36 in. cylinder, 8 ft stroke and wooden beam.
In 1794–5 Mylne recast and unified the engine-house as
the D-plan barrel-like building that survives, giving it
symmetry from the north with quadrant or curved walls.
As in his reworking of the Water House, Mylne devised a
characteristically neat solution to the extension of an
already extended building, combining tidy regularity with
a touch of grace. Smeaton's core at the centre was refitted
with a high-level cistern in its north chamber and a staircase in its south chamber, lit from the south by a new
opening, subsequently blocked. Two chimneys rose in the
south wall of the engine house extensions. The much
lower boiler-house to the south was also given symmetrically curved walls, but during the installation of a boiler
in 1796–7 it was found to be too small, and a lean-to boilerhouse evidently had to be added to the east. (ref. 49)
The replacement of the New River Company's pipes in
cast iron in the second decade of the nineteenth century
was accompanied by further upgrading of the pumping
machinery, W. C. Mylne overseeing the installation of
more powerful Boulton & Watt engines. The western
engine of 1785–6 was replaced in 1811–12 and the eastern
engine of 1794–6 in 1816–18, in both cases with singleacting parallel-motion engines with 48 in. cylinders, 8 ft
strokes and cast-iron beams. The staircase and north wall
of 1785–6 were removed, and the boiler-houses were again
enlarged for double boilers. (ref. 50) These improvements were
completed in 1818 with the removal of the staircase of
1794–6 so that Smeaton's south chamber could be reused
as a single chimney for both engines. Mylne replaced his
father's two stubby chimneystacks with a single tapering
shaft rising 110 ft from the ground (Ills 222, 223). It was
an apt moment for the raising of such a monumental
structure, which dominated the Clerkenwell skyline until
its demolition in 1954. Not only had the New River
Company brought to completion the modernization of its
water-distribution system and seen off the challenge of
rival suppliers, but it was poised to begin the most ambitious town-planning and building programme yet seen
throughout the entire district.

224. New River Head, looking west in 2005. Engine house to
rear, coal stores to right, public garden to front
Mylne oversaw another round of engine house
improvements in the 1840s, following the successful introduction by Thomas Wicksteed of a Cornish engine for the
East London Waterworks at Old Ford in 1838, and
inventions enabling the compounding of old engines to
work with high-pressure steam on the Cornish system.
Accordingly, the Boulton & Watt engines at New River
Head were altered to work expansively with new cylindrical boilers, each to provide 150 hp, one in 1845–7, the
other in 1848–9. Associated work, for which George
Mansfield & Son were the builders, included the introduction of cast-iron windows, the rebuilding in much
enlarged form of both boiler-houses, incorporating some
earlier brickwork to the south, and several additions: a
north-east porch, a staircase tower to the west, and a
seven-bay coal-store wing to the east.
All this survives relatively unaltered (Ills 221, 222, 224).
The boiler-houses and the coal-store wing each have
similar hipped roofs with light wrought-iron trusses like
those patented in France by Camille Polonceau in 1837.
The elegant and ingeniously constructed cast-iron staircase of 1848–9 was supplied by Henry and Martin De La
Garde Grissell of Regent's Canal Ironworks. These
younger brothers of Thomas Grissell, the great contractor, were leading manufacturers of structural ironwork
from c. 1841. Inside the engine house, other remnants from
this phase include some substantial cast-iron girders,
pocketed to carry the ends of floor beams in the west
engine house, and of I-section in the north chamber of
Smeaton's building, to support cisterns or condensation
tanks for preventing steam loss, under a tall iron cylinder
on an octagonal brick base behind the chimney. A third
small engine of 25 hp that appears to have been introduced
in the early 1850s may have stood below this assembly. (ref. 51)
Later in the nineteenth century a long low workshop
range was built in phases along the site perimeter north of
the engine house. The engines were again replaced in
1897–8 and 1901–3, with triple expansion machines of 65
hp by James Simpson & Co. of London, and 120 hp by Yates
& Thom Ltd of Blackburn. At this time some windows
were renewed, using fixed wrought-iron frames, and the
outer bays of the engine house appear to have been reroofed, the iron cylinder giving way to a lantern. Steam
power was replaced by electricity in 1950. With the engines
removed, a concrete floor was inserted c. 1957, the engine
house (now minus its chimney) having been refused the protection of listed-building status. This was gained in 1972,
preventing demolition, and scuppering a plan to build new
offices. The building has stood little used since, the east
boiler-house having served as a garage and then to house
generators, with alterations in 1983–5. Since 2003 the south
boiler-house has held two Siemens pumps that in serving
the London Ring Main have renewed the building's
working link with the Claremont Square reservoir. (ref. 52)

225. Metropolitan Water Board Offices, ground-floor plan as built. To left, Oak Room on first floor and (inset) its original orientation
on first floor of Water House in 1693. H. Austen Hall, architect, 1914–20
Former Metropolitan Water Board Offices
The large block of flats that has taken the name 'New
River Head' and the address No. 173 Rosebery Avenue
was built in 1915–20 as the central offices of the
Metropolitan Water Board, replacing the Water House
and displacing the Round Pond. It was in 1913, a decade
after its formation, that the board decided to build its
headquarters at New River Head. The proposal came
from Frederick Lionel Dove, who represented the
London County Council on the board. Islington-born, he
was the chairman of Dove Brothers Ltd, the eminent
Islington builders, who had undertaken several contracts
for the New River Company. His suggestion split the
board. The General Purposes Committee found little
beyond 'historical associations' in favour of siting the
headquarters outside the 'zone within which an important public body would normally erect their offices'—that
is, in or adjoining Westminster. Despite this, and higher
costs than those of a more central site, the vote went in
favour of Dove's proposal, seemingly a victory for sentiment over rationality. The inherited site, however, pro
vided scope for large premises, and it was intended that
any remaining land would be sold for development. This
goes some way towards explaining the resultant building's
orientation: there was to have been a new street along its
west side. (ref. 53)
Early in 1914 six invited architects prepared schemes,
the brief including incorporation of the Oak Room from
the Water House into the new building. The plans were
assessed by E. Guy Dawber, who favoured Herbert
Austen Hall over his competitors Brown and Barrow, T.
Edwin Cooper, Herbert O. Ellis, Edwin T. Hall, and
Henry T. Hare. (ref. 54) H. A. Hall had an appropriate track
record as an architect of town halls, and his plans were
praised by Dawber for being compact and well laid out,
'each department being kept distinct and self-contained'. (ref. 55)
Building began in July 1915 with T. W. Heath & Son as
contractors, but they proved too slow and were soon
replaced by Rice & Son of Stockwell. War brought work to
a halt between June 1916 and January 1919, and inflation
pushed up building costs. Despite economies, these rose
from the £85,000 (ref. 56) originally projected to £298,417 when
the offices opened in May 1920.
The complex comprises a quadrangle with the main
entrance on the short Hardwick Street front, and a triangle with one side aligned to Rosebery Avenue (Ill. 225).
Hall's original scheme was of Beaux-Arts character, the
main façades to have been faced in rusticated stone. But
by 1915, to reduce costs, he had substituted a more
English neo-Georgian idiom, to be built largely in red
brick over a rusticated stone lower storey. The new building rose just three storeys, with attics to the eastern ranges
(Ills 226, 230).
In 1933–6 a further attic storey was added to the eastern
ranges, with two attics to match on the western and northern ranges. Although provided for in Hall's original
scheme these do seem to detract from the building's proportions. The Portland-stone tower at the eastern angle
(Ill. 176) was added at the same time, as was a dining-room
for board members on the top floor at the north-east
corner of the main quadrangle, its cantilevered bowwindow ends offering views across London. Rice & Son
were again the builders. (ref. 57)
Internally, design was less compromised by economy.
There are several spaces in which Hall adeptly explored
the classical architectural vocabulary (Ills 227, 228). A
narrow stone-lined entrance lobby has friezes bearing
the seals of the water companies merged to form the
Metropolitan Water Board, all set over appropriately
watery Vitruvian-scroll mouldings. This leads to the spacious entrance hall that gives access to one of the two main
staircases, which have bronzed handrails and scrolled
ironwork balustrades, as well as to the former Rental
Ledger Hall, occupying the whole court within the main
quadrangle, and top-lit through an elliptically arched
ceiling. On the first floor, the former Board Room in the
south range was fitted out with a dais and concentric
benches, like a town-hall council chamber, its Ionic
columns and pilasters being a variant of the Bassae order
that had become fashionable in Edwardian London, most
prominently at County Hall, built in 1909–22.
Former Metropolitan Water Board Offices and Oak Room

226. Metropolitan Water Board Offices, perspective of south
front in 1915

227. Board Room on first floor, c. 1921

228. Rental Ledger (latterly Revenue) Hall, from south in 1937

229 (opposite). Oak Room chimneypiece

230. Former Metropolitan Water Board Offices, from Rosebery Avenue in 2006

231. Ceiling painting of William III by Henry Cooke

232. Detail of Oak Room ceiling
Conversion of the offices into 129 flats was carried out
in 1995–8 by Broadway Malyan, architects, for Berkeley
Manhattan, a consortium made up of Berkeley Homes
(Kent) Ltd, the Manhattan Loft Corporation and Kennet
Properties, a subsidiary of Thames Water. Glass conservatories were added on the roof, and the Rental Ledger
Hall, renamed the Revenue Hall, was opened up as an
enormous communal lobby. The Board Room, made
into a single voluminous flat, was redecorated for
David Dorrell to Minimalist designs by McDowell and
Benedetti. (ref. 58)
The Oak Room
The transplantation of John Grene's Oak Room of c. 1693
from the demolished Water House into the new headquarters of the Metropolitan Water Board was an early
and remarkable instance of the preservation of an historic
interior in such a context. The room had not been an incidental acquisition. On the contrary, the board had purchased it separately from the New River Company in 1904,
for £2,000, in order to prevent it being dismantled and
removed. There was thus a commitment to retaining the
Oak Room long before there were plans for a new building at New River Head. (ref. 59) This may have been at least
partly due to Charles FitzRoy Doll, an original and
actively engaged board member, who was also an architect
with antiquarian interests (see the Devil's Conduit,
below). 'Wrenaissance' taste had long appreciated the
room's quality, a watercolour view by John Crowther
having been exhibited at the Royal Academy in 1888.
Removal of the woodwork, together with the ceiling and
some later fireplaces, was entrusted to Laurence Turner.
He was able to move the ceiling in a single piece, a feat
which may have been made easier because it had been
taken down in 1868 and secured to new ceiling beams. (ref. 60)
The installation of the Oak Room on the first floor of
the new offices involved not only its re-orientation but the
loss of the panoramic views it had previously offered.
Windows could be provided on one side only, looking west
towards the engine-house. Some alteration to the ceiling
and re-organization of the panelling was involved, but
apart from that the room survives unmutilated (Ills 225,
229, 232). The decorative scheme is of the highest quality.
On the fully panelled walls are richly carved upper or
frieze panels and an overmantel with the arms of William
III framed by naturalistic ornament on fishing and
hunting themes. The carving has, inevitably, been attributed to Grinling Gibbons. However, as Gibbons's material of choice was lime-wood, it is more likely that this is
the work of another virtuoso carver, like Jonathan Maine,
who did work in the more intractable oak. (ref. 61) The enriched
plaster ceiling bears the arms of Sir Hugh Myddelton and
Grene along with numerous relief panels depicting a
variety of water subjects, ranging from angling, boats and
a swan, to dolphins, a mermaid and Neptune. Again the
identity of the craftsman is not known. Within lush
frames, gilded since at least the 1860s, there is a large
central oval oil painting by Henry Cooke, depicting William
III allegorically upheld, perhaps by Virtues (Ill. 231). (ref. 62)
Another room contains the ceiling from the east loggia
of the Water House, bearing the New River Company's
Old Testament motto 'et plui super unam ciuitatem' ('and
I caused it to rain upon one city'), (ref. 63) and the date 1693. (ref. 64)
During the Second World War the panelling was
removed by Maple & Co. for safekeeping, but it was now
impossible to remove the ceiling without cutting it up. The
central painting, restored in 1922, had to be restored again
following its reinstatement in 1945. (ref. 65)
Former Water-Testing Laboratory
The Laboratory Building at No. 177 Rosebery Avenue, on
the east side of New River Head, was built for the
Metropolitan Water Board in 1936–8 for water testing.
The board's laboratory staff of seventy had responsibility
for monitoring biological, bacteriological, chemical and
chlorination aspects of London's water. This was done in
a strikingly suave building designed by John Murray
Easton, of Stanley Hall & Easton and Robertson, a firm
beginning to specialise in medical and scientific buildings.
The builders were Walter Lawrence & Son Ltd.

233. Water-Testing Laboratory in 1938. John Murray Easton,
architect, 1936–8

234. Former Water-Testing Laboratory in 2005

235. Former laboratory staircase in 1990
Broadly functional, the building was fully attuned to
architectural fashion, the clean Modernist lines being
ideally suited to its hygienic purpose. On a steel frame, the
elevations are of brown-red Himley brick, with stone
dressings and steel windows (Ills 233–6). Like the
board's offices, the Laboratory Building faces away from
Rosebery Avenue, the entrance front here addressing
Arlington Way. The curved main range is sited away from
the road to minimise vibration, its longest convex elevation facing north to maximise the even light from this
aspect for the individual laboratories, concentrated on this
side; the presence of a basement reflects the site's former
use as a filter bed. The strong contrast between the monumental verticality of the eastern entrance block and the
sweeping horizontal lines of the main range has been
reduced by the forward extension of the latter's upper
storey in 1963–4.
For the semi-circular south end of the entrance block
John Skeaping sculpted the Metropolitan Water Board
arms in Portland stone, incorporating the New River
Company motto. Below, tall glass-brick windows light a
circular cantilevered staircase of great panache (Ill. 235).
Over the staircase is a blue plaster ceiling with a gilt incised
figure of Aquarius the water-carrier, designed by F. P.
Morton. (ref. 66) The building was converted to make 35 flats in
1997–8 by St James Homes, the joint venture that Thames
Water formed with Berkeley Homes, Geoff Beardsley &
Partners acting as architects. (ref. 67)

236. Water-Testing Laboratory, ground-floor plan as built
The Devil's Conduit
Standing just north of the former Metropolitan Water
Board offices, and within the revetment wall of the Round
Pond, is a re-sited medieval conduit head, a small, stone,
bunker-like structure known as the Devil's Conduit (Ill.
214). This was moved to New River Head in 1927, having
previously stood north-west of Queen Square in
Bloomsbury. Comprising upper and lower chambers, the
latter barrel-vaulted and ashlar-lined, it was constructed
in the fourteenth century as the conduit head or tank
at the upper end of the pipes that supplied water to
the Greyfriars monastery (later Christ's Hospital) on
Newgate Street. In 1893 Philip Norman discovered the
long disused structure behind a house and proceeded to
research and publish its history. During demolition work
for an extension to the Imperial Hotel, Russell Square,
in 1911–13, the conduit head was rescued by Charles
FitzRoy Doll, the hotel's architect. Doll had sat on the
Metropolitan Water Board since its inception, when he
had been Mayor of Holborn, as he was again in 1912–13,
and was accordingly well placed to arrange for the relic's
re-erection at New River Head. (ref. 68)
Boundary Wall, Myddelton Passage
The eastern length of purple-grey stock-brick boundary
wall on the south side of Myddelton Passage, running
from Arlington Way to the west, is the last surviving
section of New River Head's late Georgian perimeter
security wall. It was built in 1806–7 to replace a high
timber fence, and completed the enclosure of the site by
brick walls. (ref. 69) The western parts have been rebuilt in yellow
brick, perhaps in 1935, (ref. 70) and there have been smaller
yellow-brick repairs. Some of the original brickwork bears
a quantity of carved graffiti of mid-nineteenth to early
twentieth century date, discussed on page 216.