Page 1
THE BERRY HEAD CAFÉ TUNNELS REVISITED:
PREHISTORIC SOUTERRAINS OR GEORGIAN SEWAGE SYSTEM?
By Philip L. Armitage and Alan Masterson
Introduction
Among the records held in Brixham Heritage Museum’s archives, relating to past
lectures given by members of the Museum to the local history society, is a typed,
untitled two-page document summarising an illustrated lecture at Brixham
Constitutional Hall on October 27th 1976 given by the then Director of Research, Mr.
John Durston (Hon. Director of Research Brixham Museum), on the subject of a
series of tunnels discovered in March of that year at Berry Head. These tunnels
(located behind the café in the northern fort at Berry Head – see Figure 1) Durston
believed, were examples of prehistoric souterrains used by the Iron Age inhabitants of
the area as “a refuge when under attack or as a food store”; an extremely exciting
discovery as until then no known examples of such subterranean passages/chambers
were known in Devon (see Thomas 1972:75). Although Durston apparently had
hoped to keep the discovery secret from the wider public until the tunnel system had
been fully investigated, in January of the following year (1977) local and regional
newspapers were alerted to the existence of these tunnels (Herald Express 20.01.1977
and Western Morning News 21.01.1977) following damage done by vandals and
heavy rainfall, which had resulted in the collapse of the sides of a deep hole/vertical
shaft on the site of the tunnels, which had been dug by the Museum team. This
collapsed shaft was considered potentially dangerous for any visitors to Berry Head,
who were advised to keep well clear of the area. Further details of this episode are to
be found in the Brixham Museum archives, in copies of correspondence from the
Torbay Borough Engineer to the Brixham Town Clerk, expressing considerable
disquiet regarding safety aspects of the Museum’s very deep excavations on the site
of the tunnels and the subsequent collapse of the excavated shaft. Immediate remedial
measures had to undertaken by the Borough Council aimed at stabilising the
surrounding ground, and concrete rings (supported on three concrete block piers at the
bottom of the excavation) were used as a reinforced lining to the shaft, which was
then sealed by a metal manhole cover (EX in Figure 1).
Since their initial discovery made in 1976, the Berry Head tunnels have continued to
be a subject of much interest and speculation among the local Brixham people, and
the Curator at Brixham Heritage Museum (PLA) is frequently asked to explain the
age and purpose of these. It has proved difficult (until now) to answer these enquiries
with any degree of confidence owing to the limitations in the available documentary
sources, which have been restricted to the above mentioned archival records held by
Brixham Heritage Museum and a brief description published by Pye and Slater (1990:
23 – 24). What would be particularly helpful would be the original 1976/77 site
records relating to the discovery and investigation of these tunnels, but these are no
longer currently accessible to the present Museum staff owing to all contact with
Durston having ceased. Last December (2008) an opportunity arose to address this
deficiency and fill some of the gaps in our knowledge about this tunnel system. This
happened when the present authors were working with other members of Brixham
Heritage Museum’s current archaeological team in the fenced off garden area to the

Page 2
rear of the old guardhouse (Berry Head café) – digging test pits prior to the site being
handed over to Exeter Archaeology for more extensive archaeological investigation in
advance of the building of an extension to the café (part of a £1.8 million regeneration
project currently being carried out at Berry Head by Torbay Coast & Countryside
Trust Ltd.). At that time it seemed an opportune moment to explore the tunnels, and
accordingly, one of the co-authors (AM) – the thinner of the two! – volunteered to
enter and crawl along the system in order to establish/record the direction, extent, and
construction of the tunnel system, information which hopefully would shed further
light on the matters under review. In addition to the Museum’s explorations, Mr.
Mark Ledgard, the architect for Stratton & Holborow Limited, tasked with designing
the café extension, was concerned with the stability of the ground and so he
commissioned his own survey of the tunnels. It has therefore proved possible by
combining the authors’ own findings (see below) with the Stratton & Holborow’s
survey to produce a schematic plan and profile of the accessible sections of the tunnel
system behind the old guardhouse (café) – the result, drawn by Robert Rouse, Hon.
Technical illustrator for Brixham Heritage Museum, is reproduced here as Figure 1.
Supplementing the field exploratory work, co-author PLA tracked down and
contacted two former members of the Museum’s 1970s archaeological team who had
participated in the original “expeditions” through the tunnels shortly after their
discovery. One of them, Mrs. Pat Spring is still residing in Brixham, but the other,
Mr. Dennis P. Turner, now lives in Australia. Although their recollections were
somewhat hazy – not unexpected given the passage of time – information they
provided confirmed and expanded on the sketchy details already known from a letter
written by the Brixham Town Clerk, dated 26th August 1977 (Museum archives)
concerning how the tunnels were first discovered. It seems that Durston was alerted to
the possible existence of subterranean structures at Berry Head when talking to a local
builder who mentioned that in 1967, whilst putting in new drainage pipes near the
café, he had encountered what he believed was an ancient, possibly “Roman” (sic)
man-made chamber After seeking permission from the café owner, Durston’s team,
digging in the rear of the building near to the 1967 drainage works, removed what
they at first took to be a buried flagstone, but on examination was found to be a
capstone, part of the roof to a dry stone walled tunnel. Once this capstone had been
lifted, the team got their first glimpse into the tunnel system, which at that location
was completely clear of any debris. Today this original point of entry by Durston has
been retained as an access shaft covered by a wooden trapdoor (AS on the plan) –
which will eventually be inside the new café extension (now under construction) and
therefore no longer readily accessible[see updated note below]. Exploring along one
of the connecting tunnels, Durston found his passage blocked by “large boulders”,
which he was unable to remove owing to the restricted space in the tunnel. Durston
therefore crawled back along the tunnel to the café access point, and on returning to
the surface, plotted the extrapolated course of the tunnel to the blocked location. At
this point he started a surface excavation (EX on the plan – Figure 1), discovering a
circular feature (it is unclear from the available documents whether this was a natural
hollow or possibly a man-made vertical shaft?) that had been backfilled with rubble
and Napoleonic-era refuse. Removing this backfill, Durston at “considerable depth”,
entered what he interpreted as a natural water-worn cave that connected with the man-
made tunnel leading from the café he had explored earlier.

Page 3
Despite the apparent absence of any securely stratified artefacts to date the structures,
Durston firmly believed the tunnels were pre-Napoleonic and of considerable
antiquity – the few late 18th/early 19th century potsherds and military buttons found
scattered among debris on the floors of the tunnels he dismissed as later intrusive
items from the fort. It is intriguing that a letter (in Brixham Museum archives) written
by the Brixham Town Clerk (based on an interview with Durston at the time of the
collapse of the excavations) includes a mention of the finding of a single Iron Age
potsherd in the tunnels. Mr. Turner in his reply to our enquiries also recollected the
discovery of a “couple of pieces of early pottery which [as far as he recalled] were
identified as either Bronze Age or Iron Age” but he dismissed these as “almost
certainly accidental inclusions” along with the Napoleonic-era finds, which he says
were located in those places “where the cap stones had been broken through from
above”. The recent (2008/09) explorations (see below) also failed to discover any
stratified material, and only yielded a couple of clay tobacco pipe stems and three
poorly preserved (leached/abraded) fish (hake) bones scattered on the floors of the
tunnels.
A brief description of the tunnel system
Two entries were made into the tunnel system by AM from two separate points of
access, and the following accounts are based on his observations:
• The first exploration December 3rd 2008 (Figure 3)
Entry to the system was via a 1.35 m (4.4 foot) deep vertical access shaft (the size of a
modern drainage inspection chamber) located to the rear of the old guardhouse
(Figure 4). This was Durston’s original first point of entry (AS on the plan) (Grid.Ref.
SX 94332 56497). The air inside was surprisingly fresh and no residual odour was
evident as perhaps would be expected if it had been a sewage system, though this may
be explained by the tunnel not having been used for a considerable period. With the
aid of a compass the tunnel was found to run in a north westerly to south easterly
direction from beneath the guardhouse (now a café). Looking north westwards, the
tunnel at its furthest visible point curved sharply to the right (i.e. towards the NE),
“disappearing” beneath the building (marked X on the plan – Figure 1) (see also
Figure 5), but this “hidden” section beyond could not be investigated as the very
much reduced tunnel height discouraged access. In the opposite direction (south
eastwards from the access shaft), however, it proved relatively easy to crawl along the
tunnel, although this was only 0.46 m (18 inches) wide and 0.71 m (28 inches) in
height! The gently downward sloping floor was a mixture of gravel and clayey soil;
the walls were neat and well-constructed of thin limestone slabs, laid flat and keyed
without use of mortar; with the tunnel roof of rough hewn limestone capstones (see
Figure 2, section through A – A). At a point 3 m (9.8 feet) along from the access shaft
(AS), a spur/branch tunnel (also with well-constructed dry stone walls and capstones)
joined with the NW-SE tunnel, entering on the right hand side, in a slow sweeping
bend. Unfortunately this branch tunnel could not be explored beyond 3 m (9.8 feet)
owing to a blockage by rubble (stone cobs and clayey soil) up to the roof level.
Returning to the NW – SE tunnel, at a distance of 7 m (23 feet) from the access shaft
(AS), this came to a Y-junction, where it entered another tunnel. Access to the N was
blocked by debris (again stone cobs and clayey soil up to the roof level) whereas the
route extending down towards the SSW was entirely clear however, and the tunnel
along this section was slightly wider and considerably higher (1.07m) (42 inches),
suggesting this was the main tunnel in the system. Progressing along this tunnel to a

Page 4
distance extending beyond c.15 m (49.2 feet) from the Y-junction, the floor level was
seen to drop very sharply as it appeared to enter a cave. It was considered unwise to
continue the exploration given the distance covered from the access shaft and the
consequent loss of communication with other field team members at the surface, who
were (as a safety measure) monitoring progress.
• The second exploration 11th March 2009
Entry was made via the concrete lined shaft (EX on the plan) (Grid Ref. SX 94340
56476), which led into what Durston had described as a “cave” (perhaps more
accurately, a passageway through the solid bedrock). Examination of the sides of this
narrow, inverted U-shaped passageway failed to detect any evidence of shaping by
man, thus confirming Durston’s view that this had been naturally formed by water
action (see Figure 2, section through B – B). At a distance of c.12 feet (3.66m) NNE
from the modern shaft, this natural passageway sloped steeply upwards to connect
with the masonry lined (man-made) tunnel system described above – this junction is
shown in Figure 6. The builders of the tunnels apparently had utilised this natural
cave/passageway for their final section of the system, which eventually runs SSW to
emerge from below the fort’s southern garde foux into a gully in the seaward cliff
face. It was not possible to safely access the steeply descending end section of the
system (beyond Y on the plan)(Figure 7), and the presence of impenetrable blackthorn
cover prevented any external investigation of the outlet, which in any case is blocked
by rubble according to the Berry Head Manager, Mr. Nigel Smallbones, who had
examined it some years prior.
The date and purpose of the tunnel system
Having surveyed the Berry Head café tunnel system, the authors’ attention then
turned to its dating and interpretation based on the field observations, linked with
documentary research.
Considered first was the original suggestion made by Durston that the tunnels were
prehistoric souterrains. It is indeed noteworthy that in both the Berry Head system
and authenticated souterrains, the sides of the tunnels/passages comprise dry stone
walls roofed over with stone lintels/capstones, constructed in open trenches/natural
hollows and subsequently buried in made up, levelled ground. In the case of Berry
Head this made up ground consists of limestone rubble, possibly originating from the
rock removed during the digging of the fort’s defensive ditch and/or provided from
the adjacent quarry. Notwithstanding this similarity in method of construction, there
are however two features which stand out to distinguish the two systems: in the
authenticated souterrains the tunnels are markedly wider and higher and these
generally terminate in chambers (many of beehive shape with corbelled roofing) –
similar chambers are apparently entirely absent at Berry Head.
An alternative explanation for the tunnels had been put forward by Pye and Slater in
their 1990 report on the Berry Head fortification (Pye and Slater 1990:24) – “the
system relates to the fort buildings and [therefore] probably represents drains and/or
sewers of Napoleonic date”. After exploring the tunnels, and based on many years of
working in the construction industry (including as a local authority drainage technical
officer) co-author AM came to the conclusion this interpretation seemed highly
plausible. If correct, the tunnels presumably had been installed c.1794/95 during the

Page 5
initial construction phase of the fort, or slightly later, between c.1798 – 1802, when
the guardhouse had been built.
In order to provide supporting evidence for the alternative explanation, an internet
(Google) search was carried out by the other co-author (PLA) to locate any published
examples of drainage/sewage systems associated with 18th/19th century forts. An
initial reference found to a “souterrain” associated with the Mount Pleasant Redoubt
(Plymouth) proved to be a disappointingly false lead. On checking the published
paper by Pye (1992) it was revealed this structure was in fact an underground barrel
vaulted corridor providing access between the redoubt’s central blockhouse and the
main magazine situated under the south west corner of the rampart – nothing at all to
do with sewage/drainage! After a series of other false leads, reference was eventually
found to underground drainage/sewage systems at Fort Henry, Kingston, Ontario
(Canada). An email enquiry directed to Mr. Ron Ridley, Curator at Fort Henry,
resulted in one of his colleagues, Mr. Bob Garcia, Historian, Ontario Service Centre
(Parks Canada) kindly providing a pdf copy of his report on these systems (Garcia
2003). Reading through the report by Garcia of the description of the underground dry
stone walled drains/sewers servicing the privies at Fort Henry it was immediately
apparent there were striking similarities in the construction and dimensions to the
tunnels at Berry Head – despite the fact that the Canadian system was of later date
(1839 – 1850s, during the British phase in Fort Henry’s history).
Conclusion and discussion
Based on the recent reappraisal, the authors are of the opinion that the Berry Head
interconnecting tunnels to the rear of the café (old guard house) once formed part of
the fort’s sewage/drainage system and therefore are of Georgian date and not of
prehistoric origin as originally proposed. There are however unresolved aspects
related to the system that merit further research, including the following:
It is unclear exactly where the privies were located in the fort and how these
connected with the sewage/drainage system. As mentioned in a previous CBA SW
Journal article (Armitage 2007:20) there is no evidence for any sanitary facilities
within the barrack buildings.
It is also unclear whether there was provision for periodically flushing waste through
the sewage tunnels and into the sea, using water channelled from the slate-covered
roofs of the barrack buildings and stored in a “tank” (see Evans 1986:14) for this
purpose. At Fort Henry (Canada) there was such a cleansing system utilising water
collected from grates in the parade ground - which was improved with the installation
(in 1858) of a reservoir/storage tank allowing controlled discharge from time-to-time
of the collected water by opening a sluice gate; the subsequent flow of water flushed
out the waste into the adjacent bay (Navy Bay).
An archaeological watching brief by Exeter Archaeology is currently underway at the
café site during the building of the extension and it is to be hoped additional details
relating to the Georgian-period tunnels may come to light, which will contribute
further to our understanding of the sewage system.

Page 6
ACKNOWLEDGEMENTS
The authors wish to express their gratitude to the following people who provided
invaluable input to the project: Nigel Smallbones, Manager, Berry Head National
Nature Reserve (Torbay Coast & Countryside Trust); Jane Hirst, Project Manager,
“On the Edge” Berry Head Regeneration Project (Torbay Coast & Countryside Trust);
Hal Bishop, Archaeological Planning Officer (Torbay Council); Ron Ridley, Curator,
Fort Henry National Historic Site of Canada (Ontario Ministry of Tourism); Bob
Garcia, Historian, Parks Canada (Ontario Service Centre); Mark Ledgard, Architect
(Stratton & Holborow Limited); Bob Rouse, Hon. Technical Illustrator (Brixham
Heritage Museum); Gerry Perkins, Field Research Team member (Brixham Heritage
Museum); John Maule, Hon. Photo Archivist (Brixham Heritage Museum); Andrew
Passmore (Exeter Archaeology); and sincere thanks are extended to the two ex-
members of the Brixham Museum Archaeological 1970s Team, Pat Spring and
Dennis Turner.
References
Armitage, P. L. 2007 “Mean diminutive wooden sheds … comprised of thin planks.
The 1803 infantry barracks at Berry Head (Torbay)”. CBA SW Journal No. 20: 17 –
21.
Evans, D. 1986 The History of the Berry Head Fortifications. Report produced for
the Bridge Agency.
Garcia, B. 2003 Underground Drainage at Fort Henry Volume 1. Parks Canada.
Pye, A. R. and Slater, W. D. 1990 Berry Head Fort, Brixham An Archaeological
Survey. Exeter Museums Archaeological Field Unit Report No. 90.10.
Pye, A. 1992 “An archaeological survey of Mount Pleasant Redoubt, Plymouth”.
Devon Archaeological Society Proceedings No. 50: 137 – 161.
Thomas, C. 1972 “Souterrains in the Sea Province: a note. The Iron Age in the Irish
Sea Province”. CBA Research Report 9: 75 – 78.
Westerbury, K. 1977 “Vandals damage Iron Age tunnels at Berry Head”. Herald
Express (Brixham) 20/01/77.
Western Morning News 1977 “Vandals and rain ruin Iron Age find”. Western
Morning News 21/01/77, p.7.
This article was originally published in Council for British Archaeology South-West
Journal No. 23 – June, 2009: 15 – 21. The website version replaces the original
black/white photographs with colour pictures. The access shaft today (2011) now is
located within the new café and protected by a glass viewing panel.

Page 7
CAPTIONS TO ILLUSTRATIONS
Figure 1: Location map, plan and profile of the tunnel system inside Berry Head Fort
No.3. (Drawn by Robert Rouse).
Figure 2: Sections through A – A and B – B shown in Figure 1.
(Drawn by Robert Rouse).
Figure 3: Co-author Alan Masterson preparing to descend into the tunnel system via
the modern access shaft to the rear of the café (formerly the old guardhouse).
(Photo: Philip L. Armitage).
Figure 4: View down into the access shaft. (Photo: Philip L. Armitage).
Figure 5: View northwest along the tunnel to where it turns sharply to the right (far
distance) to continue beneath the café (old guardhouse) – point X on the
plan (Figure 1).
(Photo: Alan Masterson).
Figure 6: View from the 1976/77 excavated shaft (EX on the plan) towards the
junction where the natural cave connects with the man-made tunnel system.
(Photo: Alan Masterson).
Figure 7: View southwards from inside the 1976/77 excavated shaft (EX on the plan).
(Photo: Alan Masterson).


Figure 1


Figure 2


Figure 3


Figure 4


Figure 5


Figure 6


Figure 7