Evidence for extensive ice cover on the Isles of Scilly

Abstract

Previous studies of the Quaternary sediments on the Isles of Scilly have suggested that the ice of the Late Devensian Irish Sea Glacier impinged upon the north coasts of the islands but that it did not extend much further south. There is also disagreement in the literature about whether any earlier glacial episode affected the islands, although a number of researchers have noted the presence of erratic cobbles in pre-Devensian raised beaches. The new field research, based in part upon examinations of exposures that might not have been available to earlier researchers, shows that erratic cobbles and pebbles are common in raised beaches and in early and middle Devensian brecciated slope deposits around all of the island coasts. The most parsimonious explanation is that they have come from disaggregated glacial deposits dating from at least one glacial episode (Anglian?) during which the islands were completely inundated by ice. Furthermore, coherent diamictons similar to those found on the north coasts of the islands, and in western and southern Pembrokeshire, are also found on the coasts of St Mary’s and St Agnes islands, indicating that the Devensian ice cover was more extensive than previously suggested. It appears that ice from the Celtic Sea pressed into the archipelago from the north-west and west, with lobes fingering into sounds and straits which are currently below sea-level. The diamictons, rich in striated and faceted erratic cobbles and pebbles from many different lithologies, do not appear to be primary tills; they are suggestive of an ice-margin environment in which disaggregation and redistribution of glacial and glaciofluvial sediments has occurred. The new interpretation of Late Glacial Maximum ice extent is consistent with other recent work which places the limit of the Irish Sea / Celtic Sea ice lobe around 250 km south-west of the Isles of Scilly, near the Celtic Sea shelf edge.

Full text

Journal of the Quaternary Research Association
The QRA home age on the world wide web can be found at: http://www.qra.org.uk
The Quaternary Research Association is an organisation comprising archaeologists, botanists, civil
engineers, geographers, geologists, soil scientists, zoologists and others interested in research into
the problems of the Quaternary. The majority of members reside in Great Britain, but membership
also extends to most European countries, North America, Africa, Asia and Australasia. Membership
(currently c. 1,200) is open to all interested in the objectives of the Association.
EDITOR:
Dr A. Stone
Geography, School of Environment, Education and Development, The University of
Manchester, Arthur Lewis Building, Oxford Road, M13 9PL
(e-mail: abi.stone@manchester.ac.uk)
Brian John, 2018. EVIDENCE FOR EXTENSIVE ICE COVER ON THE ISLES OF
SCILLY. Quaternary Newsletter Vol. 146, October 2018, pp 3-27.
ABSTRACT
Previous studies of the Quaternary sediments on the Isles of Scilly have suggested that the ice of the Late Devensian
Irish Sea Glacier impinged upon the north coasts of the islands but that it did not extend much further south. There is
also disagreement in the literature about whether any earlier glacial episode affected the islands, although a number of
researchers have noted the presence of erratic cobbles in pre-Devensian raised beaches. The new field research, based
in part upon examinations of exposures that might not have been available to earlier researchers, shows that erratic
cobbles and pebbles are common in raised beaches and in early and middle Devensian brecciated slope deposits around
all of the island coasts. The most parsimonious explanation is that they have come from disaggregated glacial deposits
dating from at least one glacial episode (Anglian?) during which the islands were completely inundated by ice.
Furthermore, coherent diamictons similar to those found on the north coasts of the islands, and in western and southern
Pembrokeshire, are also found on the coasts of St Mary’s and St Agnes islands, indicating that the Devensian ice cover
was more extensive than previously suggested. It appears that ice from the Celtic Sea pressed into the archipelago from
the north-west and west, with lobes fingering into sounds and straits which are currently below sea-level. The
diamictons, rich in striated and faceted erratic cobbles and pebbles from many different lithologies, do not appear to be
primary tills; they are suggestive of an ice-margin environment in which disaggregation and redistribution of glacial and
glaciofluvial sediments has occurred. The new interpretation of Late Glacial Maximum ice extent is consistent with
other recent work which places the limit of the Irish Sea / Celtic Sea ice lobe around 250 km south-west of the Isles of
Scilly, near the Celtic Sea shelf edge.
https://www.qra.org.uk/uploads/documents/QN146-Oct_2018_Colour.pdf

Quaternary Newsletter Vol. 146 October 2018 3
ARTICLE
EVIDENCE FOR EXTENSIVE ICE COVER ON THE ISLES OF
SCILLY
Brian John
Introduction
The Isles of Scilly have featured prominently in the literature because they support
the southernmost coherent glacial deposits on land in the British Isles (Scourse,
1991, 1998; Hiemstra et al., 2006; McCarroll et al., 2010). These deposits,
concentrated on the northern shores of St Martin’s, Tresco and Bryher, were once
thought to be of pre-Devensian age (Mitchell and Orme, 1967), but radiocarbon,
OSL and TCN dating and examinations of organic content have subsequently
demonstrated convincingly that they were laid down at the margin of the Irish Sea
Glacier near the LGM (Scourse, 1991; Smedley et al, 2017) (Figure 1). Following
work by Praeg et al. (2015), Clark et al. (2018), Glasser et al. (2018) and Small
et al. (2018) have suggested that the LGM occurred around 27,000 years ago,
and that the outer limit of the Irish Sea Ice Stream lay about 250 km SW of the
Isles of Scilly, at the Celtic Sea shelf edge .
The basic Quaternary stratigraphy of the islands has been described by Scourse
(1991, 1998), modified after Mitchell and Orme (1967). The oldest deposits
described thus far are the raised beaches which are widespread on all of the
island coasts. Above these are soliflucted granite breccias full of locally derived
angular clasts. These slope deposits (in some locations stratified, as in West Wales)
are suggestive of a cold climate, although it is by no means certain that the full
thickness (over 4m in some places) was laid down under periglacial or permafrost
conditions. Organic remains in layers of finer sediments suggest that at times the
climate was suitable for a tundra grassland to thrive. Within the recorded glacial
limit on the northern islands the lower slope deposits are intermittently overlain
by diamictons related to a short-lived glacial incursion, and then by an upper slope
deposit of granite debris which occupies the same stratigraphic position as the
“upper head” in West Wales (Campbell and Bowen, 1989). There are also many
exposures of a sandy reddish loess-like material. There are a number of different
facies, affected by a range of processes, and Scourse (1991) suggests that at the
time of original deposition there was a barren landscape and an ice edge not far
away. Some of the fine-grained windblown material has been redistributed as
colluvial “drapes” following deglaciation, and its uppermost layers incorporate
the modern soil horizon.

4 Quaternary Newsletter Vol. 146 October 2018
Figure 1. The previously proposed maximum extent of Devensian ice in the Isles
of Scilly, defined by the limit of the Hell Bay Gravel. Source: Scourse (1991
and 1998).

Quaternary Newsletter Vol. 146 October 2018 5
Observations from the Isles of Scilly, April 2016
During the course of a week’s visit to the islands in April 2016 it was possible
to walk most of the coastline of the islands of St Mary’s, St Martin’s, St Agnes,
Bryher and Tresco. Visits were also made to Gugh and White Island. Unfortunately
it was not possible to land on Samson, Annet, the Eastern Isles or the outer islets.
Some of the exposures visible in April 2016 appeared quite fresh following winter
storms, and some have changed appearance since Scourse conducted his detailed
research on the islands in 1981-85 and since the QRS field meeting of 2006.
Some of the exposures noted in this short paper may not previously have been
investigated. It is emphasised that the following text is based upon a set of initial
field observations, made under less than ideal conditions and interpreted in the light
of the author’s own field experience and following constructive comments from
referees and the editor of this journal. There was no time for sample collection or
meticulous stratigraphic recording or sediment analysis, and no attempt has been
made to fit the author’s stratigraphy into the lithostratigraphic units of Scourse
(1991, 1998) or to define till exposures into categories related to mechanisms of
formation (cf. Evans, 2017). However, the article is offered in the hope that it
will provide food for thought, raise alternative hypotheses, and encourage further
detailed studies of the sites described.
1. Lowest slope deposits
In most of the literature about the glaciation and geomorphology of the Isles
of Scilly, the raised beach is referred to as the lowest and oldest deposit in the
Quaternary stratigraphic sequence, resting directly on bedrock (Hiemstra et al.,
2006). However, two sites revealing lower deposits have now been discovered.
Near Browarth Point on St Agnes (SV878088), raised beach gravels, cobbles and
boulders are solidly cemented and stained with manganese oxide, in a layer up to
50 cm thick. They lie on deeply rotted granite and “grus” which is also cemented
with manganese and iron oxides. Most of the gravelly particles are less than 1cm
in diameter, and larger bedrock fragments appear to have rotted more or less in
situ, with only a little downslope movement. More varied slope deposits are found
under the raised beach near
Dutchman’s Carn on the west side of Peninnis Head, St Mary’s (SV909095). At
this location (Figure 2) there are around 2 m (base not seen) of cemented coarse
granite breccias made mostly of grus but with silty and sandy layers or lenses.
There are signs of crude stratification, and many clast long axes are aligned with
the bedrock slope.
2. The raised beach
Raised beaches are ubiquitous, and it is almost impossible to walk along the
coastline of any one of the islands without seeing a cliff exposure somewhere.

6 Quaternary Newsletter Vol. 146 October 2018
Figure 2. Coarse blocky slope deposit with a matrix of grus beneath the raised
beach on the west side of Peninnis Head, St Mary’s.
They are unlikely to be misinterpreted, since the great majority of exposures lie
adjacent to modern cobble beaches with very similar characteristics. Many of
the locations were mapped by Mitchell and Orme (1967). The type locality is at
Watermill Cove, on the east side of St Mary’s Island (SV924123), where Scourse
(1991) refers to it as the lowest (and oldest) of the Quaternary sediments in the
archipelago. It contains unlithified sand beds and is overlain by a poorly-sorted
breccia made of gravel, sand and angular granite clasts of many different sizes.
There are other exposures of the raised beach on the west side of St Mary’s,
between Porthloo and Carn Morval (SV907116). Some of these exposures are
well below current HWM (high water mark), where they are distinguished from
modern beach sediments by the fact that they are solidly cemented and blackened
by manganese oxide. Closer to Carn Morval (SV906118) these cemented materials
can be seen at the base of the sediment cliff, full of well-rounded erratic and local
pebbles and cobbles. Scourse (1991, p 28) also noticed that some raised beach
exposures contain “a considerable proportion of erratic clasts.” In some sections
the gravels and pebble bands are mixed with faceted erratics in a silt and clay
matrix and are sealed beneath brecciated slope deposits.

Quaternary Newsletter Vol. 146 October 2018 7
There is a cemented raised beach resting on a rock platform at SV871148 on
the south side of Gweal Hill on Bryher. At Porth Seal, St Martin’s (SV918166)
there is disputed evidence for two raised beaches, separated by a coarse breccia.
At present, the most accessible exposures of the raised beach are to be seen in the
bay of Porth Killier (SV880086), at the northern tip of St Agnes Island (Figure
3). The cobbles exposed in cliff faces are for the most part well-rounded. In some
parts of the beach they are packed together, and in others held in a sandy and
silty matrix which may have been filtered into position at some stage following
beach formation (Scourse, 2006). Scourse described the boulders and cobbles
as being vertically-inclined; but this characteristic was not observed in 2016.
Rather, it was observed that in many exposures the arrangement of long axes
was somewhat chaotic, suggestive of post-depositional disturbance. The beach
here is partly or solidly cemented with iron oxide and manganese oxide – this is
a common feature, but there are some places where the beach is uncemented and
friable enough to extract cobbles with ease from exposures. In all of the examples
seen on St Agnes, there are rounded cobbles and pebbles of sandstone, chert,
mudstone and some volcanics in the beach.
Figure 3. Raised beach exposure in Porth Killier, St Agnes. Here the beach
materials are overlain by up to 2m of windblown deposits and colluvium.

8 Quaternary Newsletter Vol. 146 October 2018
At Porth Killier there are not many sediments above the raised beach, because
there are no adjacent steep gradients which could supply slope materials over
a prolonged period of time. However, in places there is an overlying layer of
rich brown sand and silt (with signs of layering) and a modern organic-rich
soil horizon. In some exposures at the northern end of St Agnes, thin clay-rich
diamictons containing faceted non-local cobbles lie directly on the raised beach
or on a capping layer of sandy loess and colluvium. This has apparently not been
noticed by earlier investigators.
Most of the raised beach deposits are found at around extreme high-water mark,
but some are found down to mid-tide mark, beneath the cobbles and boulders of
the present-day beach. No raised beaches (apart from the Garrison Boulder Bed
on St Mary’s) were found more than 5 m above HWM.
3. Other slope deposits
Pseudo-stratified slope deposits are seen most clearly on the shore of Porthloo
Bay (SV908115) in a cliff that can be followed along the beach at low tide for
several hundred metres (Scourse, 1991) (Figure 4). The material, consisting for
the most part of a granite breccia of angular clasts derived from an old buried
Figure 4. Coarse granite “solifliction breccia” with pseudo-stratification, exposed
in the cliffs at Porthloo on the island of St Mary’s.

Quaternary Newsletter Vol. 146 October 2018 9
cliffline, is up to 4 m thick. There are a number of discontinuous and sometimes
contorted horizons of fine-grained materials. In places the lowest part of the
cliff face consists of foliated or laminated sandy and gravelly layers which are
relatively free of large clasts. There are also lateral variations along the cliff face,
with lenses of finer debris adjacent to coarser brecciated materials.
Rockfall and slope deposits were referred to as as “main head” and “upper head”
by Barrow (1906) and by some later authors, as “ram” by Knight (2015), and as
“Porthloo Breccia” and “Bread and Cheese Breccia” by Scourse (1998). These
terms can be confusing, but they are useful in indicating that there are, in places,
two lithostratigraphic units of different ages. As Scourse (1998) has pointed
out, in some locations in the north of the islands the two stratigraphic units are
separated by layers of other deposits, whereas in the south, in most sites, there
appears to have been little or no discontinuity between the accumulation of the
one and then the other.
At Bread and Cheese Cove (SV 940159), a site on St Martin’s which was not
easy to examine in 2016, Scourse (1991) referred to a lower layer of “Porthloo
solifluction breccia” which appears to be free of erratics, capped by a layer of
Scilly Till with associated glaciofluvial gravels, and above that a layer of “Bread
and Cheese solifluction breccia” which does contain occasional erratics.
In the cliffs on the east side of Porthcressa (SV908098) some of the brecciated slope
deposits appear to have been churned both above and below an undulating layer
of silt and sand about 10 cm thick (Figure 5). On the east coast of Bryher, south of
Norrard and The Bar (SV881153), there are many exposures of reddish gravelly
solifluction materials in the low cliffs. These are generally less than 2m thick in
exposures, and contain no clear stratigraphic breaks, although in some sections
the base seems to consist of rotted granite more or less in situ with soliflucted
materials above, grading upwards into more sandy materials and then a darker
organic-rich horizon near the surface. Pebbles with non-local lithology are seen
on the beaches and occasionally in the cliff sections – red and pink sandstones,
limestone, grey shales and mudstones, and some schists. There are also many
different types of granite derived from local outcrops. Reddish gravelly slope
deposits are also exposed in the cliffs of Rushy Bay (SV876142), at the southern
end of Bryher. The gravels are partly cemented by iron and manganese oxides,
and blocks of this material have fallen from the cliff face.
4. Glacial deposits and landforms
The northern tip of the island of Tresco lies just within the recorded limit of the
Devensian glaciation of the British Isles (Scourse, 1998). A diamicton with highly
variable clast sizes, shapes and lithologies is exposed in the cliffs to the north of
Cromwell’s Castle (SV881161), and there are also patches of erratic-rich gravels
to the south of Gimble Porth (SV891153).

10 Quaternary Newsletter Vol. 146 October 2018
Erratics and patches of similar material occur in cliff exposures even further to the
south in the strait between Bryher and Tresco. Exposures of related diamictons
have now been exposed in many other locations within the accepted ice limit,
generally at the coast in low cliffs with granite breccia and other deposits, but
sometimes at the ground surface inland.
On St Martin’s the exposure of clay-rich diamicton at Bread and Cheese Cove
has been described by Mitchell and Orme (1967), Scourse (1991, 1998) and
Hiemstra et al. (2006). It contains abundant erratics and complex structures
(Figure 6). Other diamictons with erratics can also be seen in the core of White
Island Bar (SV924171), and in the spectacular cleft called Chad Girt (SV925173).
Scourse (1998) noted that no in situ “Scilly till” was exposed in the Chad Girt
cleft at the time of his own field research, and suggested that the glacial debris
which he examined had been soliflucted or redeposited. However, recent erosion
along the cleft has revealed a c 2 m thick layer of massive clay-rich diamicton
similar in appearance to the Irish Sea till of west Wales (Figure 7). From a visual
assessment, the matrix appears to contain a higher percentage of clay and a lower
percentage of gravel than the deposit at Bread and Cheese Cove. Unfortunately
Figure 5. “Churned” slope deposits of breccia in a sandy and gravelly matrix
beneath silts and sands with a high organic content, exposed on the east side of
Porthcressa on St Mary’s. There are no apparent fossil frost wedges.

Quaternary Newsletter Vol. 146 October 2018 11
the exposure was too wet to pick up any signs of deformation or re-deposition.
The current stratigraphy at the site is as follows:
1. Cemented raised beach -- 50 cm
2. Lower slope deposits, cemented and heavily stained -- 2m
3. Massive clay-rich diamicton with erratics -- up to c 2 m
4. Upper slope deposits -- c 50 cm
On Bryher there are many exposures of diamictons with erratics, on the western
and northern flanks of Gweal Hill (SV873150), at the southern and northern
ends of Popplestones Bay (SV874152), and in the west-facing cliffs of Hell
Bay (SV876157). The material is associated with perched blocks at the western
extremity of Gweal Hill (870148), and there are signs of depositional ridges nearby.
Inland, on the high plateaux of Bryher, Tresco and St Martins diamictons are found
up to an altitude of 40m, with a scatter of erratics in a gravelly and sandy matrix.
There are some gravelly diamictons on the slopes of Hell Bay. On Tregarthen
Hill (SV886164) there are some sandy silts and also thin patches of undisturbed
Figure 6. The 2016 exposure of clay-rich gravelly diamicton (Scilly till) at
Bread and Cheese Cove, containing abundant erratics and traces of sand lenses
and shearing.

12 Quaternary Newsletter Vol. 146 October 2018
clay-rich deposits containing abundant erratic pebbles and cobbles of up to about
20 cm in length. Some of these show clear striae. The erratic suite is described
by Scourse (1991) as typical of all of the glaciallyderived deposits on the islands.
Identifications of erratics by JR Hawkes (1991) include assorted sandstones
(some pink and red pebbles), shale, greywacke, quartzite, chert, conglomerate,
acid igneous rock and siltstone.
Other coastal exposures of erratic-rich diamicton have now been found to the
south of the putative Devensian ice limit. On St Mary’s there are traces of this
material near Carn Morval (SV906118), close to the site described in detail by
Scourse (1991). Here there has been substantial recent coastal erosion. The deposit
is less than 1 m thick, and was seen in 2016 in close association with a cemented
raised beach and granite breccia. However, it is distinct from those deposits in
that both rounded cobbles and a wide range of faceted erratics are set in a matrix
of silt and clay. It is separated from the underlying raised beach by an erosional
contact and a band of relatively clast-free silt and clay up to 15 cm thick.
Figure 7. Massive clay-rich diamicton exposed in the cleft at Chad Girt on White
Island. The matrix of this material is similar to that seen at Bread and Cheese
Cove on St Martin’s, but there is a greater percentage of clay and silt.

Quaternary Newsletter Vol. 146 October 2018 13
On St Agnes there are many exposures of similar diamictons. Most of these are
on the north and west coasts. The material is nowhere more than 50 cm thick,
and it is difficult to interpret because it is in some cases not particularly clay-rich,
but it consists largely of reworked raised beach materials mixed with grus and
occasional faceted non-local cobbles. To the south of Carnew Point (SV875079)
a similar deposit (Figure 8) lies above 1m of light brown silts showing some
bedding structures, and beneath 1m of gravelly granite breccia capped by 1m of
darker silts and sands at the ground surface.
At White Par (SV875078), near Long Point on the west side of St Agnes Island,
a similar diamicton lies on top of 2.5 m of grus with granite clasts in a gravelly
granular matrix. The silty and sandy material incorporates faceted pebbles, and
also contains granite breccia blocks and some raised beach materials. This is the
site stratigraphy:
Figure 8. Diamicton exposure on the west coast of St Agnes, near Carnew
Point. Note the wide range of far-travelled erratics. The deposit may have been
redistributed following original emplacement at or near a glacier front.

14 Quaternary Newsletter Vol. 146 October 2018
1. Base not seen – masked by modern beach.
2. About 2.5 m of weathered granite debris / breccia with large angular granite
clasts.
3. No sharp junction – layers grade imperceptibly one into the other
4. 50 cm of mixed diamicton with faceted pebbles of grits, sandstones, shales and
unknown volcanics. Some cobbles of raised beach origin are incorporated.
5. Sharp junction
6. 1m of light brown stratified silts and sands – colluvium
7. Top – ~1 m of dark organic-rich silts and sands topped with modern soil.
The White Par diamicton with erratics is very similar to that seen at Popplestones
and in parts of the long cliff exposure in Bread and Cheese Cove.
On Badplace Hill, Bryher (SV875163), there are rock surfaces that appear to
be ice-scoured, intermittently covered with thin turf. There are also a number
of constructional features which may be associated with glaciation, including
a substantial ridge connecting White Island to St Martin’s, the Pernagie Bar,
an apparent boulder moraine on Shipman Head, and surface irregularities on
the north-western flank of Bryher, running out to Popplestones and Gweal Hill
(SV873149). These features are described by Hiemstra et al. (2006) and by
McCarroll (2016). Other ridges described in Bread and Cheese Cove and on the
northern tip of St Martin’s, beneath Rabbit Rocks, are in the view of the present
author not convincing since they could be related to bedrock outcrops.
5. Glaciouvial deposits
Coherent glaciofluvial deposits are not widespread in the Isles of Scilly, and
apart from thin traces none were seen on the 2016 visit which is the subject of
this paper. However, they have been described in a few locations by Mitchell
and Orme (1967) and by Scourse (1998), particularly near the northern coasts.
6. Aeolian deposits
In many coastal exposures there are layers or “drapes” of fine-grained reddish
or dark-coloured deposits that are relatively stone-free (Figure 9). They occur in
a number of stratigraphic positions. They are in places made of coarse sand, and
sometimes they are silty, with or without an organic content. At Porth Seal there
are both sand and silt layers beneath brecciated slope deposits. At Popplestones
sandy deposits with included clasts overlie stratified gravels. These deposits
contain occasional complex flow structures that pass over or under incorporated
clasts. The foxy red sandloess, sometimes more than 1 m thick, occurs near the
top of many Quaternary sediment exposures, resting on granite breccia or raised
beach deposits and incorporating a black or dark brown soil layer.

Quaternary Newsletter Vol. 146 October 2018 15
7. Other glacial traces
Faceted, sub-angular and sub-rounded pebbles and cobbles are found in the
Porthloo Breccia and in all other deposits in coastal exposures well beyond the
putative Devensian ice limit. For example, there are many erratics in the raised
beaches around the head of Porth Killier Bay (SV880085), near the northern tip
of St Agnes Isle. There are also small faceted erratics in the lowest breccia and
grus on both sides of Beady Pool (SV884074). There are yet more erratics on
the coasts of Wingletang Down, scattered through raised beach and solifluction
materials. Near the southern tip of the island (SV884072) small faceted pebbles
of red sandstone, grey shale and bluish rhyolite have been found in the granite
breccia, 2m beneath its upper surface. These observations accord with Scourse
(1991), who pointed out, there are erratics in most of the raised beaches of the
islands.
On the east side of Gugh Island a striated cobble of chert has been found deeply
embedded in weathered granite breccia or grus, in a position that could not have
been related to the Devensian glaciation (SV892082) (Figure 10). On the east
Figure 9. A “drape” of sandy and silty reddish material overlying erratic-rich
diamicton at Popplestones on Bryher. In places there are apparent flow structures.
There is a high organic content in the darker surface layers / soil horizon.

16 Quaternary Newsletter Vol. 146 October 2018
coast of Bryher, on the shore of New Grimsby Harbour, an erratic cobble of red
sandstone was found deeply embedded in gravel breccia or head (SV881151). On
the east coast of Bryher, south of Norrard and The Bar (SV881153) the pebble
lithology on the beach and in cliff sections (red and pink sandstones, limestone,
grey shales, mudstones and some schists) are erratics. At Tregear’s Porth, east
of Watermill Cove on St Mary’s island (SV926122) there are abundant red and
pink sandstone boulders which are not of local origin.
Figure 10. Striated and fractured erratic cobble found in granite breccia and grus,
on the east side of Gugh Island. There are other small erratics in the vicinity.
Interpretation
1. Lowest slope deposits
The Watermill Sands and Gravels (incorporating the raised beach)were referred
to by Scourse (1998, p 258) as “the basal stratigraphic unit of the Quaternary
succession on the islands”. However, the presence in some localities of grus
and slope deposits with signs of mobilisation beneath the Watermill Sands and
Gravels unit should not be a contentious issue. Indeed, it would be a surprise if
across the whole archipelago, with a long coastline, the raised beach was always
found to rest upon coherent bedrock. The most parsimonious explanation of the

Quaternary Newsletter Vol. 146 October 2018 17
deposits beneath the raised beaches on St Mary’s and St Agnes is that they are
slope deposits and rockfall debris accumulated from weathered granite debris
prior to the period of beach formation. No evidence has been seen which might
suggest the presence of permafrost, and no geochronological data is available
for this debris.
2. The raised beach
The field observations of raised beach deposits made here are that most are found
in the extreme high-water mark and some in the mid-tide mark. This accords with
the observation made by Scourse (1991; 1998), who furthermore demonstrated
that most of the accessible raised beach and associated deposits are of littoral
origin, and that they must date from the last interglacial. The importance to
geomorphology of the Watermill Cove (SV924123) lithostratigraphic unit, observed
by Scourse (1991) lies not so much in its diagnostic beach characteristics as in
the pollen content and dating of the overlying sediments (Scourse, 1991). It may
have formed at a time of relatively high sea-level in the archipelago, somewhat
above present MSL (mean sea level).
In all of the examples seen on St Agnes, there are rounded cobbles and pebbles
of sandstone, chert, mudstone and some volcanics in the beach, suggestive
of the incorporation of pre-existing foreign materials. Scourse (1991, p 441)
noted that some exposures “contain a wide range or erratic clasts.” Mitchell and
Orme (1967) claimed that the raised beach deposits are occasionally strongly
cryoturbated, and indeed there are some exposures where cobbles have their long
axes vertically inclined, suggestive of post-depositional disturbance either by
periglacial processes or by overriding ice. Mitchell and Orme also suggested that
there are TWO raised beaches of different ages in the islands, since some beaches
are cemented and stained with a black manganese oxide crust, while others are
“fresh” in appearance. With respect to the suggestion, Scourse (1991) and other
later researchers have disagreed and have argued forcefully that there is just one
ancient beach, which dates from the Last Interglacial. This shows considerable
variation in its altitude above OD and in its internal characteristics; stained layers
sometimes cross stratigraphic junctions, suggesting that they are more influenced
by groundwater conditions than age. In some places (as at Battery) the raised
beach has been disaggregated and redistributed by slope processes and by glacial
action. The same is true of the raised beach at Port Killier. Observations by the
present author support the view that there is but one raised beach.
3. Other slope deposits
The crude stratification and clast alignment in these deposit types both suggest
downslope movement, and the label of “solifluction breccia” is appropriate.
Bands of iron oxide precipitate or hardpan appear to coincide with old surfaces
in a sequence of accumulating slope deposits. Most of the angular fragments

18 Quaternary Newsletter Vol. 146 October 2018
of varied lithologies appear to have come from granite or related igneous rock
outcrops seen in the vicinity, notably from buried clifflines or coastal slopes.
The structural characteristics of some deposits seen in the cliffs of Porthloo Bay
suggest that they have been affected by occasional slumps, debris flows and other
erosional episodes. However, in those locations beyond the putative Devensian
ice edge the process of breccia accumulation seems to have continued without a
break, and where there are signs of stratigraphic disruption the unconformities
appear to be quite localised. The type locality at Porthloo is described by Scourse
(1991 and 1998). There is uncertainty about the significance of signs of churning
in some slope and sandloess deposits, as at Porthcressa (Figure 5). Cryoturbation
in solidly frozen ground might have occurred, but rapid mass movement might
also have occurred in places within a mobile active layer. In general, there seems
no good reason for the use of the term “periglacial” in connection with the bulk
of these deposits. However, Hiemstra et al. (2006) show, on the basis of the
micromorphology and clast fabrics of the slope deposits in Bread and Cheese
Cove, that at least some of them were periglacial gelifluctates.
4. Glacial deposits
The interpretation of assorted diamictons, exposed north of Cromwell’s Castle
(SV881161), the south of Gimble Porth (SV891153) and in the strait between
Bryher and Tresco, as glacial deposits within the putative Devensian glacial limit
appears to be well founded (cf. Hiemstra et al., 2006). The diagnostic characteristics
of the till at Bread and Cheese Cove and elsewhere include striations and facets
on pebble and cobble surfaces, a wide range of represented lithologies, and a
clay-rich matrix with gravelly, sandy and silty layers. On the basis of detailed
studies of fabrics, particle sizes and internal structures in the Scilly Till, Hiemstra
et al. (2006) interpreted it as “a mixture of glacitectonized and cannibalized
glacilacustrine and marine deposits as well as primary subglacial tills.” They
suggested that the Scilly Till at Bread and Cheese Cove was the only exposure of
coherent Devensian till on the Isles of Scilly. However, the massive till at Chad
Girt (Figure 7) deserves further study. It is quite possible that further exposures
will be discovered in the future.
The thin diamictons outside the putative glacial limit also merit detailed
examination, since they are similar in appearance to those along the coast from the
type section in Bread and Cheese Cove, above the raised beach and in association
with slope deposits. They may be in situ, and may not be; this does not much
matter in the context of the present exercise, and even if they are “paraglacially
remobilised” or tectonized, they must first have been carried in and deposited
somewhere in the neighbourhood of the locations where they are currently
exposed. As for erratic sources, the purple, pink and red sandstones and siltstones
are reminiscent of certain Devonian and Cambrian sedimentary rocks found in
Devon, Pembrokeshire and Ireland. Scourse (1991) provided accurate lithology

Quaternary Newsletter Vol. 146 October 2018 19
percentages for pebble groups in the three lithostratigraphic units deemed to be
of glacial origin and identified possible offshore sources to the north. Without
further petrographic and geochemical investigation more accurate provenancing
is not possible here. Smedley et al. (2017)referred to a quartzite erratic on Tresco
that may have come from Anglesey or the east coast of Ireland. The largest glacial
erratic found on the islands is a 10 tonne block of olivine basalt found by Dr JR
Hawkes on Great Crebawethan Rock (SV832070), about 3 km south-west of St
Agnes. There is no way of knowing how long it might have been there; it may
have nothing to do with the Devensian ice incursion. Since 1990 it has not been
seen, but it must still be somewhere in the vicinity. Scourse (1998) and Smedley
et al. (2017) suggest that the distribution of eroded or rounded tors on the islands,
particularly on the northern and eastern coasts, provides a guide to the position of
the Devensian ice limit; but there is so much variation in tor morphology that in
the view of the present author much closer examination of the evidence is required.
The clay-rich diamictons described above, containing faceted erratic cobbles and
pebbles and material derived from nearby raised beaches, found on the islands
of St Mary’s, St Agnes and Gough, lie several kilometres beyond the putative
Devensian ice limit of other researchers. In the northern isles the distribution of
glacial deposits on Tregarthen Hill and Castle Down to the east and Shipman
Head Down to the west make it possible that a lobe of ice occupied the low land
between Bryher and Tresco. The distribution of deposits on the coasts of St Mary’s
and St Agnes may also suggest that a lobe of ice pushed in from the west. Future
investigations of Annet and Samson might reveal whether those islands have been
affected by Devensian ice flowing from the west or north-west. The diamicton at
White Par is interpreted here as the most southerly exposure of coherent glacial
deposits on land in the British Isles.
5. Glaciouvial deposits
No new glaciofluvial deposits were described during the field observations in
this study. Scourse (1998) argued that pulses of glaciofluvial outwash deposition
occurred at Battery (SV887165) on the north coast of Tresco at the end of the
Devensian glaciation. He suggested that some of the sandy gravels and thin
beds of cobbles are primary meltwater deposits which were later affected by
aeolian and periglacial processes. Dates of around 19,000 yrs BP from the top of
the sequence appear to confirm this (Smedley et al., 2017). These authors also
reported OSL dates of around 25 ka for the lower part of the Battery outwash
sequence, representing the earliest part of deglaciation.
6. Aeolian deposits
The sandy and silty drapes observed in the field are interpreted to be of aeolian
origin, which is consistent with Barrow (1906) and Scourse (1991, 1998), who
argued that most of the drapes are aeolian accumulations formed by strong

20 Quaternary Newsletter Vol. 146 October 2018
winds in environments where extensive areas of glaciofluvial and old sea-floor
deposits were not fixed by vegetation. Most of the exposures are classified as
belonging to the Old Man Sandloess lithostratigraphic unit. Suitable windblown
and relatively arid environments must have been extensive within and beyond the
Scilly archipelago during and after the Late Devensian glacial episode, given that
sea-level was depressed by at least 100m at the time. There may have been some
redeposition under moist conditions, and indeed the term “colluvium”might be
more appropriate than the term “sandloess” in some of the locations investigated.
Scourse (2006) considered that some of the sandloess deposits exposed in cliff
sections have accumulated in standing water or have been fluvially redistributed.
7. Pre-Devensian glacial deposits
There is ongoing uncertainty about older glaciations in the Isles of Scilly, and
the observation of erratics in most of the raised beach units observed here, and
by Scourse (1991, p437) support the idea that there was a more extensive pre-
Devensian glaciation that affected the islands: “........the late Devensian event was
probably not the first glacial event to have influenced the Islands because erratics
are widespread in some exposures of the Watermill Sands and Gravels, but this
earlier event is quite distinct from the late Devensian advance”. However, Hiemstra
et al. (2006, p308) stated that “the evidence suggests that an Irish Sea Glacier
has only reached so far south on one occasion.” The observations presented in
this paper suggest that Scourse was correct in suggesting an extensive ice cover
on more than one occasion. The faceted, sub-angular and sub-rounded pebbles
and cobbles found in the Porthloo Breccia, and in all other deposits in coastal
exposures, are well beyond the putative Devensian ice limit. On the east side of
Gugh Island the striated cobble of chert embedded in grus is in a position that
could not have been related to the Devensian glaciation (SV892082) as currently
defined (Figure 1). There is no reason to think that any of the boulders or cobbles
identified as erratics were imported within the last few centuries as ship’s ballast,
on the basis of their locations observed here and in the literature. The Wingletang
Down erratics are most likely derived from disaggregated earlier glacial deposits,
since it is difficult to envisage any other processes by which they could have been
deposited upslope of their present locations. Near the southern tip of St Agnes
erratic pebbles are set so low in the stratigraphic sequence, embedded in Porthloo
breccias, that they could not have had anything to do with the Devensian glacial
incursion. If it should be argued that they were introduced at or near the LGM, an
explanation would have to be found for the lack of older Devensian /Ipswichian
materials beneath them.
Discussion
It is now possible to suggest a revised Late Quaternary relative chronology for the
Isles of Scilly. The lowest slope deposits and grus layers may be millions of years

Quaternary Newsletter Vol. 146 October 2018 21
old, but they are most likely to date from a pre-Ipswichian cold phase, given that
there seems to be no unconformity or deeply weathered horizon between them
and the overlying raised beach, which Scourse (1991; 1998) has demonstrated to
date to the Last Interglacial (Ipswichian). Remarkably little is known about the
landscape impacts of cold phases prior to this in the British Isles, but whether
or not it was cold enough for an extensive glaciation in the Celtic Sea arena, it
would be reasonable to suggest, in tune with climatic reconstructions (Polyak et
al., 2018) that prior to the last interglacial, periglacial conditions prevailed for
thousands of years.
On the matter of early glacial deposits, McCarroll (2016) suggested that there are
no erratics on the island of St Mary’s, and that to the south of the Devensian ice
limit “erratics are completely absent.” However, glacially-derived erratics were
found in all of the coherent Quaternary deposits on all of the islands visited by this
author in April 2016. Without chronological data it is not possible to definitively
place the timing of these early glacial deposits.
Scourse (1991) argued convincingly that there was no Wolstonian glaciation of
the Isles of Scilly, the adjacent coast of Cornwall, or the Celtic Sea to the west.
Gibbard and Clark (2011) and Rolfe et al. (2012), on the basis of evidence from
other parts of South-West England and Wales, have suggested that the pre-
Devensian Irish Sea Glacier which affected the Celtic Sea coasts can be dated
to the MIS12 or Anglian glacial episode c 450,000 years ago. The abundance
of erratics on east-facing coasts in the Isles of Scilly (observed in this study)
means it is reasonable to assume that this same glacier overwhelmed the Isles of
Scilly, leaving no part ice-free and terminating some way to the south and east.
If so, it may have been more extensive and no less powerful than its Devensian
successor (Praeg et al., 2015; Glasser et al., 2018). However, doubts about the
age of the “Greatest British Glaciation” (GBG) persist, and without chronological
evidence this remains a hypothesis. There are also doubts about the dimensions
and glaciological characteristics of this early version of the Irish Sea Glacier.
One must also consider the possibility that the scattered erratics in the lower
deposits in the stratigraphic sequence have been ice-rafted and incorporated into
slope deposits and raised beaches; in other words, they might not be indicative
of a complete glacial cover of the archipelago. In the view of the present author
this would be a complex explanation. It is difficult to conceive of an episode
of ice-rafting onto all the shores of the Isles of Scilly at a time when sea-level
in the islands was higher than that of the present day. Conversely, sea-levels at
times of ice rafting of glacial debris are most likely to have been more than 100m
lower than the sea-level of today. A full glacial cover, leaving behind scattered
glacial deposits which were later eroded away and reworked, is perhaps the most
parsimonious explanation for the features described above. It is suggested that
the ice crossed the islands from the north-west towards the south-east; if the
cliffs of Cornwall acted as a barrier to ice progress, and if ice pressed into the

22 Quaternary Newsletter Vol. 146 October 2018
approaches to the English Channel, ice flow must have been perpendicular to the
ice edge, and this greatly reduces the possibility of ice movement from north to
south (Patton et al., 2017).
Turning to the age and stratigraphic position of the raised beach, the most
parsimonious explanation for its position in the regional sedimentary stratigraphy
is that it still lies where it was formed in the Ipswichian interglacial (cf. Scourse,
1991). There appears to be no evidence that the raised beaches resting on slope
materials at Browarth Point and Dutchman’s Carn have been soliflucted down from
a higher altitude on the cliff faces, thus coming to rest on top of slope deposits
that are younger than the original raised beaches. These must be interglacial storm
beaches thrown up onto pre-existing sediments. This is a situation commonly
encountered in modern storm beach environments, for example at Porth Morran
on St Martin’s and at West Angle, Newgale and Aber-mawr in Pembrokeshire
(Campbell and Bowen, 1989). Indeed, at Aber-mawr a storm beach ridge has
migrated landwards in recent centuries across unconsolidated “submerged forest”
deposits of peat with roots, fallen branches and tree trunks, without destroying
them (Bell, 2007). On exposed headlands in the archipelago occasional clusters of
well-rounded cobbles and even boulders are found in and on top of slope deposits,
and even on grassy banks in association with damaged turf, having been thrown
up during extreme storm events.
The observations recorded here support the view of Scourse (1991) and other
authors that the brecciated slope deposits found above the raised beach around
most of the coasts of the islands represent a prolonged cold period in which cold
or periglacial conditions predominated, as in the coastal sections of Pembrokeshire
and Gower (John, 1973; Campbell and Bowen, 1989; Hiemstra et al., 2008). There
are some signs of cryoturbation, suggesting the intermittent presence of permafrost.
Outside the putative Devensian glacial limit the accumulation of these deposits
(designated Porthloo Breccia) continued. Inside the limit it is possible to identify
two horizons of brecciated solifluction material separated in some locations by
glacial and glaciofluvial materials; as in West Wales, the upper horizon contains
more non-local components, including faceted and striated erratics (Campbell
and Bowen, 1989).
The fieldwork results reported in this paper do not support an Early Devensian
glaciation on the lines proposed by Rolfe et al. (2012), since the Porthloo solifluction
breccias occupy the stratigraphic position in which any glacial deposits of this
age might be expected. The most parsimonious explanation of the diamictons
described (for example at Chad Girt, Carn Morval, Popplestones and White Par)
is that they are deposits of Late Devensian till, related in age and origin to the
“stratotype” of the Scilly Till at Bread and Cheese Cove but displaying no signs
of glacitectonic structures. However, diamictons are not necessarily diagnostic of
a till. It is possible that the Carn Morval diamicton, for example, was deposited
as a stony colluvium carried downslope across an eroding raised beach surface,

Quaternary Newsletter Vol. 146 October 2018 23
although the overlying unit of a clay-rich sediment containing erratics of many
sizes, shapes and lithologies is indicative of a glacial incursion. It is suggested
that glacier ice has pressed onshore in the vicinity of the carn and has overridden
and incorporated a range of pre-existing deposits. The erratic-rich diamicton on
the north and west coasts of St Agnes may either be a till, or a glaciotectonite,
formed as ice moved across an old coastline, overriding and incorporating large
quantities of old storm beach and shingle beach materials and also introducing
far-travelled erratics in the process. Further descriptions of the sediments are
required to investigate this further.
It is not at all unusual for several broadly contemporaneous types of till to be
exposed within a relatively small area, as in Pembrokeshire and Llyn (John,
1970b; McCarroll, 2001). The association of faceted and striated erratic pebbles
of various sizes, beach cobbles, and clay-rich matrix, combined with a consistent
stratigraphic position above the undisturbed raised beach and beneath upper
solifluction breccia, wind-blown and colluvial sediments and modern soil, is
best explained, in south Pembrokeshire and the perhaps the Isles of Scilly, as a
consequence of a glacial incursion.
Some of the recorded Devensian Scilly tills appear to have moved into depressions
or gaps between granite outcrops in an ice edge and ice wastage environment.
There is no reason at this stage to suggest that all of the till was “redistributed”
hundreds or thousands of years after the ending of the glacial incursion across
the coastline, as suggested by Bowen (1984). Hiemstra et al. (2006) refer to the
sequence of events in the islands as glacitectonization of sea-floor sediments
pushed onshore, followed by paraglacial redistribution and then by periglacial
disaggregation, all within the same recent glacial episode. The evidence collected
in April 2016 broadly supports this interpretation, with the proviso that not all of
the Devensian tills appear to have been glacitectonized.
The glaciofluvial materials described by Scourse (1991) at Battery and elsewhere
are associated with ice wastage on the islands of Tresco and Bryher, and are
normally assigned to the “Tregarthen Gravels”. However, on Tregarthen Hill
and elsewhere the sands and gravels appear to grade laterally into patches of
clay-rich and gravelly till with faceted and striated erratics, suggestive of near-
contemporaneous deposition. Since the glacier ice that affected the islands must
have had a crenulated margin, occupying the bays and sounds and leaving some
higher adjacent land ice-free, it is suggested that most true fluvioglacial deposits
and landforms will have been concentrated in areas now submerged by the sea.
May (1980) suggests that glaciofluvial material could have later been disaggregated
and redistributed as beach and sea floor deposits.
A suggested ice limit for the Late Devensian incursion by the Irish Sea Glacier
would require only modest adjustments to the ice front positions portrayed by
Scourse (1989) and Hiemstra et al. (2006) on the north coasts of the islands.
However, the glaciogenic deposits described on the west coast of St Mary’s

24 Quaternary Newsletter Vol. 146 October 2018
and St Agnes suggests that Devensian glacier ice pressed into St Mary’s Road
and came into contact with the west coast of St Mary’s and the north coast of St
Agnes (Figure 11).
A Late Devensian ice incursion from the west would not be surprising given the
till fabric evidence presented by Hiemstra et al. (2006) that suggested that the
ice which affected the northern coasts of the islands came not from the north
but from the northwest, flowing south-eastwards. It now seems probable from
the work of Praeg et al. (2015) and Clark et al. (2018) that the ice of the BIIS
(British Irish Ice Sheet) Celtic Sea lobe extended all the way to the shelf edge.
There would have been no topographic constraint on the eastern edge of this lobe,
and since ice in this situation must have flowed perpendicular to the ice edge, it
is entirely reasonable to postulate that it pressed into the basin between Bryher
and St Agnes, flowing broadly from west to east and impinging on some of the
present-day coasts in the process, as supported by field observations in this study.
Figure 11. Modified glacial limits for the Devensian glaciation of the Isles of
Scilly, based largely upon evidence presented in this paper. The red line represents
the limit of the Hell Bay Gravels, taken by Scourse (1991) and others to represent
the maximum extent of Devensian ice. The black line shows a revised Devensian
limit, incorporating ice incursion from the west into the southern islands.

Quaternary Newsletter Vol. 146 October 2018 25
It is noteworthy that there is a very close match between the Quaternary stratigraphy
of the Isles of Scilly and that of western and southern Pembrokeshire (John, 1970,
1973; Campbell and Bowen, 1989). The raised beach is ubiquitous around the
coastline of both areas. Above it, the Devensian sequence of thick lower brecciated
slope deposit, then till and related glaciofluvial deposits, and then a thin upper
brecciated slope deposits, is replicated almost exactly. The greatest difference
between Pembrokeshire and Scilly is that in the former, coastal exposures reveal
occasional great thicknesses of till (sometimes in excess of 5 m), while in the
latter the till is thin and patchy. In Pembrokeshire there is also greater relief, and
accumulations of slope materials are, as expected, greater than in the low-lying
islands of Scilly. Also, in Pembrokeshire many of the slope deposits show signs
of permafrost active layer activity, whereas there appear to be no fossil ice wedges
and related structures in Scilly, and just a few signs of cryoturbation.
Conclusions
From field observations made in 2016 it is suggested that two glacial episodes
(and maybe more) have affected the Isles of Scilly. During the earlier of these,
postulated as the Anglian or MIS-12 glaciation, the ice of the BIIS inundated the
whole of the archipelago and terminated some distance to the east and south. Raised
beaches occur in many different locations, at different altitudes and displaying
different lithological characteristics. However, their relationship to other deposits
is consistent, supporting the idea that they are of Last Interglacial (Ipswichian)
age. The cold episode which followed culminated in a Late Devensian glacial
incursion which affected only the northern and western coasts of the islands. The
glacier ice must have flowed from northwest to south-east or from west to east.
The new observations reported in this paper, in part made possible by new coastal
exposure conditions, contribute to a long and fascinating debate among Quaternary
researchers. It is hoped that these observations may help to elucidate the glacial
history of the Isles of Scilly and encourage more detailed field observations,
sedimentological and geochronological analysis.
References
Barrow, G. (1906). The Geology of the Isles of Scilly. Memoirs of the Geological
Survey of Great Britain. HMSO, London. 418 pp.
Bell, M. (2007). Prehistoric Coastal Communities: The Mesolithic in Western
Britain, CBA Research Report 149, 2 pp.
Campbell, S., Bowen, D.Q. (1989). Quaternary of Wales. Geological Conservation
Review, NCC, 237 pp.
Clark, C. D., Ely, J. C, Greenwood, S. L., Hughes, A. L. C., Meehan, R., Barr,
I. D., Bateman, M. D., Bradwell, T., Doole, J., Evans, D. J. A., Monteys, X.,

26 Quaternary Newsletter Vol. 146 October 2018
Pellicer, X. M., Sheehy, M. (2018). BRITICE Glacial Map, version 2: a map
and GIS database of glacial landforms of the last British–Irish Ice Sheet. Boreas,
Vol. 47, pp 11–27.
Evans, D.J.A. (2017) Till: a glacial process sedimentology. Wiley, 400 pp.
Gibbard, P.L., Clark, C.D. (2011). Pleistocene glaciation limits in Great Britain. Ch
7 in Quaternary Glaciations: extent and chronology. Developments in Quaternary
Science, Vol 15, pp 75-93.
Glasser, N.F., Hambrey, M.J., Davies, B.J., Gheorghiu, D.M., Balfour, J., Smedley,
R.K., Duller, G.A.T. (2018). Late Devensian deglaciation of south-west Wales
from luminescence and cosmogenic isotope dating. Journal of Quaternary Science
doi.org/10.1002/jqs.3061
Hawkes, J.R. (1991) Appendix 2 in Scourse (1991) Late Pleistocene Stratigraphy
and Palaeobotany of the Isles of Scilly. Philosophical Transactions of the Royal
Society B, 334 (1271), 405-448.
Hiemstra, J.F., Evans, D.J.A., Scourse, J.D., McCarroll, D., Furze, M.F.A., Rhodes,
E. (2006). New evidence for a grounded Irish Sea glaciation of the Isles of Scilly,
UK. Quaternary Science Reviews, 25, 299-309.
Hiemstra, J. F., Rijsdijk, K. F., Shakesby, R. A. and McCarroll, D. (2008).
Reinterpreting Rotherslade, Gower Peninsula: implications for Last Glacial ice
limits and Quaternary stratigraphy of the British Isles. Journal of Quaternary
Science, 24, 399-410.
John, B.S. (1970a). Pembrokeshire, in Lewis, C.A. (ed) The Glaciations of Wales
and Adjoining Regions, Longman, 378 pp.
John, B.S. (1970b). The Pleistocene drift succession at Porth-clais, Pembrokeshire,
Geological Magazine 107 (5), pp 439-457.
John, B.S. (1973). Vistulian Periglacial Phenomena in South-West Wales, Biuletun
Peryglacjalny 22, pp 185-212.
Knight, P. (2015). Glaciers in the Isles of Scilly. http://www.petergknight.com/
scilly/glaciers.html
May, V.J. (1980). Chapter 8: Sand spits and tombolos – GCR site reports: Isles
of Scilly (GCR ID: 2185). Volume 28: Coastal Geomorphology of Great Britain,
p 2731.
McCarroll, D. (2001). Deglaciation of the Irish Sea Basin: a critique of the
glaciomarine hypothesis. Journal of Quaternary Science, 16 (5), 393-404.
McCarroll, D., Stone, J.O., Ballantyne, K., Scourse, J.D., Fifiled, K., Evans,
D.J.A., Hiemstra, J.F. (2010). Exposure-age constraints on the extent, timing
and rate of retreat of the last Irish Sea ice stream. Quaternary Science Reviews,
29, 1844-1852.

Quaternary Newsletter Vol. 146 October 2018 27
McCarroll, D. (2016). Trimline Trauma: The Wider Implications of a Paradigm
Shift in Recognising and Interpreting Glacial Limits. Scottish Geographical
Journal, 132(2), 130-139.
Mitchell, G.F.,Orme, A.R. (1967). The Pleistocene deposits of the Isles of Scilly.
Quarterly Journal of the Geological Society of London, 123, 59– 92.
Patton, H., Hubbard, A., Andreassen, K., Auriac, A., Whitehouse, P.L., Stroeven,
A.P., Shackleton, C., Winsborrow, M., Hyeman, J., Hall, A. M. (2017). Deglaciation
of the Eurasian ice sheet complex. Quaternary Science Reviews, 169, 148-172.
Polyak, V.J., Onac, B. P., Fornós, J. J., Hay, C., Asmerom, Y., Dorale, J. A., Ginés,
J., Tuccimei, P., Ginés, A. (2018). A highly resolved record of relative sea level
in the western Mediterranean Sea during the last interglacial period. Nature
Geoscience, 2018; DOI: 10.1038/s41561-018-0222-5
Praeg, D., McCaroon, S., Dove, D., Ó Cofaigh, C., Scott, G., Montey, X., Facchin,
L., Romeo, R., Coxon, P. (2015). Ice sheet extension to the Celtic Sea shelf edge
at the Last Glacial Maximum. Quaternary Science Reviews, 111, 107-112.
Rolfe, C.J. ,Hughes, P. D., Fenton, C. R., Schnabel, C., Xu, S., Brown, A. G.
(2012). Paired 10Be and 26Al exposure ages from Lundy: new evidence for the
extent and timing of Devensian glaciation in the southern British Isles. Quaternary
Science Reviews, 43, 61-73.
Scourse, J.D. (1991). Late Pleistocene Stratigraphy and Palaeobotany of the Isles
of Scilly. Philosophical Transactions of the Royal Society B, 334 (1271), 405-448.
Scourse, J.D. (1998). The Quaternary History of the Isles of Scilly (Chapter 8 in
Quaternary of South-West England, edited by S. Campbell et al.) Joint Nature
Conservation Committee; Springer Science, 249-269.
Scourse, J.D. (ed) (2006). The Isles of Scilly: Field Guide. Quaternary Research
Association, London, 2006, 180 pp.
Smedley, R.K., Scourse, J.D., Small, D., Hiemstra, J.F., Duller, G.A.T., Bateman,
M.D., Burke, M.J., Chiverrell, R. C., Clark, C.D., Fabel, D., Gheorghiu, D. M.,
McCarroll, D., Medialdea, A., Xu, S. (2017). New age constraints for the limit of
the British–Irish Ice Sheet on the Isles of Scilly. Journal of Quaternary Science,
32 (1), 48–62.
Brian John
Trefelin, Cilgwyn,
Newport,
Pembrokeshire SA42 0QN.
brianjohn4@me.com

NOTE: In the published version of this paper the bottom part of this map was inadvertently
chopped off -- removing the locations of sites on St Agnes and Gugh which are crucial for
the interpretations made in the paper. Please therefore ignore the map on page 24 of the
article and refer to this one instead.
Figure 11. Modified glacial limits for the Devensian glaciation of the Isles of Scilly,
based largely upon evidence presented in this paper. The red line represents the limit
of the Hell Bay Gravels, taken by Scourse (1991) and others to represent the maximum
extent of Devensian ice. The black line shows a revised Devensian limit, incorporating
ice incursion from the west into the southern islands.