Lithostratigraphy of Miocene–Recent, alkaline volcanic fields in the Antarctic Peninsula and eastern Ellsworth Land

Abstract

Miocene–Recent alkaline volcanic rocks form numerous outcrops scattered widely throughout the Antarctic Peninsula and eastern Ellsworth Land. They occur mainly as short-lived (typically 1–2 million years) monogenetic volcanic fields but include a large outcrop area in northern Antarctic Peninsula which includes several substantial polygenetic shield volcanoes that were erupted over a 10 million year period (the James Ross Island Volcanic Group (JRIVG)). As a whole, the outcrops are of considerable importance for our understanding of the kinematic, petrological and palaeoenvironmental evolution of the region during the late Cenozoic. Until now, there has been no formal stratigraphical framework for the volcanism. Knowledge of the polygenetic JRIVG is still relatively poor, whereas a unifying lithostratigraphy is now possible for the monogenetic volcanic fields. For the latter, two new volcanic groups and twelve formations are defined, together with descriptions of the type sections. The volcanic fields (both polygenetic and monogenetic) vary in area from c. 1 to 4500 km2, and aeromagnetic data suggest that one may exceed 7 000 km2. The rocks are divisible into two contrasting petrological ‘series’, comprising basanites–phonotephrites and alkali basalts–tholeiites. The JRIVG is dominated by alkali basalts–tholeiites but also contains rare basanites, and phonotephrite–tephriphonolite compositions occur in minor pegmatitic segregations in sills. By contrast, in the monogenetic volcanic fields, basanites–phonotephrites generally form the older outcrops (mainly 15–5.4 Ma) and alkali basalts–tholeiites the younger outcrops (4(?)–<1 Ma). Throughout the region, erupted volumes of alkali basalts–tholeiites were an order of magnitude greater, at least, than those of basanite–phonotephrite compositions. Interpretation of the lithofacies indicates varied Miocene–Recent palaeoenvironments, including eruption and deposition in a marine setting, and beneath Alpine valley glaciers and ice sheets. Former ice sheets several hundred metres thick, and fluctuating ice surface elevations, which were generally higher during the eruptive periods than at present, can also be demonstrated.

Full text

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
Antarctic Science
71
(3):
362-378
(1999)
0
British Antarctic Survey Printed in the United Kingdom
Lithostratigraphy of Miocene-Recent, alkaline volcanic fields in
the Antarctic Peninsula and eastern Ellsworth Land
J.L.
SMELLIE
British Antarctic Survey, Natural Environment Research Council,
High
Cross, Madingley Road, Cambridge CB3
OET,
UK
Abstract:
Miocene-Recent alkaline volcanic rocks form numerous outcrops scattered widely throughout the
Antarctic Peninsula and eastern Ellsworth Land. They occur mainly as short-lived (typically 1-2 million years)
monogenetic volcanic fields but include
a
large outcrop area in northern Antarctic Peninsula which includes
several substantial polygenetic shield volcanoes that were erupted overa
10
million year period (the James Ross
Island Volcanic Group (JRIVG)). As
a
whole, the outcrops are
of
considerable importance for our
understanding of the kinematic, petrological and palaeoenvironmental evolution of the region during the late
Cenozoic. Until now, there has been no formal stratigraphical framework for the volcanism. Knowledge of the
polygenetic JRIVG is still relatively poor, whereas
a
unifying lithostratigraphy is now possible for the
monogenetic volcanic fields. For the latter, two new volcanic groups and twelve formations are defined,
together with descriptions of the type sections. The volcanic fields (both polygenetic and monogenetic) vary
in area from
c.
1
to
4500
km2,
and aeromagnetic data suggest that one may exceed 7
000
km'. The rocks are
divisible into two contrasting petrological 'series', comprising
basanites-phonotephrites
and alkali basalts-
tholeiites. The JRIVG is dominated by alkali basalts-tholeiites but
also
contains rare basanites, and
phonotephrite-tephriphonolite
compositions occur
in
minor pegmatitic segregations in sills. By contrast,
in
the
monogenetic volcanic fields,
basanites-phonotephrites
generally form the older outcrops (mainly
15-5.4
Ma)
and alkali basalts-tholeiites the younger outcrops
(4(?)+1
Ma). Throughout the region, erupted volumes of
alkali basalts-tholeiites were an order
of
magnitude greater, at least, than those
of
basanite-phonotephrite
compositions. Interpretation of the lithofacies indicates varied Miocene-Recent palaeoenvironments, including
eruption and deposition in amarine setting, and beneath Alpine valley glaciers and ice sheets. Former ice sheets
several hundred metres thick, and fluctuating ice surface elevations, which were generally higher during the
eruptive periods than at present, can
also
be demonstrated.
Received
20
January
1998, accepted 29 March 1999
Key
words:
alkaline volcanism, Antarctic Peninsula, lithostratigraphy, monogenetic, palaeoenvironment,
polygcnetic, volcanic field
Introduction
The Antarctic Peninsula region has been
a
major continental
consuming plate margin for most of Mesozoic and Cenozoic
time, with
a
complicated history of plate interactions (e.g.
Grunow
1993,
Storey et
al.
1996, McCarron
&
Larter 1998,
McCarron
&
Smellie 1998). Subduction ceased progressively
northward during the Cenozoic,
as
offset sections of
a
spreading
centre collided with the trench (Barker 1982) except
at
the
South Shetland trench, where it continued at
a
very slow rate
(Larter 1991). Each section of downgoing slab is thought to
have become detached along the former spreading axis,
resulting in the development of "no-slab windows" and the
generati on of small-volume asthenospheric melts with alkaline
compositions (Holeetal. 1991a, 1993,1995, Hole&Saunders
1996). These were erupted throughout the Antarctic Peninsula,
forming small, short-lived monogenetic volcanic fields (Fig.
I
;
Smellie 1987, Smellie
et
al.
1988,
Hole 1988, 1990a).
Conversely, where subduction continued
(as
at the South
Shetland trench), slab window development did not occur and
plume-related alkaline magmas were erupted in the back-arc
region, forming large polygenetic shield volcanoes with
multiple satellite centres and
a
prolonged history of eruptions
extending over
10
million years (James
Ross
Island area;
Smellie 1987, Sykes
1988a,
Hole
et
al.
1995).
Rapid
accc~
to the surface for the alkaline magmas was facilitated
by
thc
presence ofblockfaults (Smellie 1987). Although the faulting
is related to regional extension, Holeetal. (1995)demonstratcd
that major extension has probably not occurred. The alkaline
volcanism
was
coincident with glacial conditions in the region
and many of the volcanic outcrops in the region show evidence
forinteraction withglacialmeltwater (e.g. Smellie
etal.
19'33,
Smellie
&
Skilling
1994,
Smellie
&
Hole 1997).
This paper summarizes the principal results of recentresearch
into the alkaline volcanic fields
in
the Antarctic Peninsula
and
eastern Ellsworth Land and uses the information to erect thc
first rigorous lithostratigraphy
.
Outcrops in the South Shetland
Islands, included
in
aprevious review ofalkaline volcanism
in
the region (Smellie
et
al.
1988) are excluded here; despitc
some alkaline characteristics (principally high sodium
contents), they are otherwisc "normal" calc-alkaline
and
362

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY OF MONOGENETIC VOLCANIC FIELDS 363
tholeiitic basalts probably related to slow subduction at the
South Shetland trench (unpublished information of the author).
With the exception of the James Ross Island Volcanic Group
(JRIVG), all of the stratigraphical names used in this paper are
new. The outcrops discussed are an important source of
information on the kinematic, petrological and
palaeoenvironmental evolution of the Antarctic Peninsula
region during the late Cenozoic.
Systematic descriptions
of
the volcanic lithostratigraphy
Classical lithostratigraphical procedures, which are based
predominantly on sedimentary formations, can be difficult to
apply to volcanic outcrops (e.g. Ricci
etal.
1993). According
to Whittaker
etal.
(199
I),
defined formations "should have or
should once have had physical continuity". Sedimentary
formations are also normally limited to a single sedimentary
basin and names should not normally be extended to adjacent
basins. These conditions
are
violated by many volcanic fields
in which large and small volumes of erupted magma are
clustered around multiple discrete volcanic centres, which
may
or
may not overlap laterally. Terrestrial volcanic fields
may also form no obvious part of a sedimentary basin. In this
paper it is proposed that volcanic fields can be regarded as the
volcanic equivalent of sedimentary basins. Thus, the presence
of discrete volcanic fields. separated geographically and not
overlapping, can be used as aprincipal criterion for identifying
and delimiting volcanic formations. Thereafter, normal rules
for distinguishing lithostratigraphical units are applied; for
example, presence of bounding unconformities and lithological
distinctiveness within each volcanic field (cf. Whittaker
et al.
1991). Using these criteria, the group status of the volcanic
and sedimentary JRIVG is reaffirmed, and two new groups
containing twelve new formations are identified.
Distinguishing characteristics
of
the different rock units are
summarised in Table I. The division into rock groups broadly
follows LeMasurier
&
Thomson (1990), who divided the
Antarctic Peninsula region into volcanic provinces. Outcrops
ofthe two new groups (Seal Nunataks and Bellingshausen Sea
groups) are separated geographically by
>
500
km and they are
situated
in
the tectonically contrasting fore- and back-arc
regions of the former Antarctic Peninsula magmatic arc. The
Merrick Mountains outcrops are unusual for the region
in
being situated within the Mesozoic arc terrain. They were
grouped by LeMasurier
&
Thomson (1990) within the
Bellingshausen Sea Volcanic Province and that association is
continued here. Compositional names
of
the lavas are after
Le Bas
et
al.
(1986).
James
Ross
Island Volcanic Group
Nameandhistory:
Originally called the Ross Island Formation
by Andersson
(1906),
whoconducted theearliest studies. The
name was changed to James
Ross
Island Volcanics by Adie
(1953) and, finally, James
Ross
Island Volcanic Group (JRIVG)
Fig.
1.
Sketch map of the Antarctic Peninsula
and
eastern
Ellsworth Land, showing the location and distribution
of
the
alkaline volcanic fields discussed
in
this paper.
by Nelson (1975). However, constituent formations were not
defined. Reconnaissance studies of parts of the sequence were
made by British geologists (Bibby 1966, Baker
et
al.
1977,
and unpublished Falklands Islands Dependencies Survey
reports by Croft 1947 and Stoneley 1952). The only
geographically comprehensive investigation was by Nelson
(1
979, who also published
a
geological map of the entire
group. The most detailed study, including lithofacies
descriptions, isotopic (K-Ar) ages and
a
large database
of
chemical analyses, was by Sykes (1988a,1988b), although
that work was confined to outcrops mainly in northern James
Ross Island.
Distribution:
James Ross Island, Vega Island and several
smaller islands nearby
in
Prince GustavChannel and Antarctic
Sound (Fig. 2), southern Dundee Island and Paulet Island.
Tabarin Peninsula and Cain and Abel nunataks (Graham
Land). A related dyke also crops out on Snow Hill Island
(Massabie
&
Morelli 1977, Malagnino 1978).
Subdivision:
Nelson (1975) divided the JRIVG into five
volcanic phases, each dominated by
a
lava unit and cogenetic
palagonite-altered breccia, and associated sedimentary strata.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
Table
I.
Summary lithostratigraphy. principal characteristics,
age
and Interpretation of Miocene-Recent alkaline volcanic fields
in
the Antarcti Peninsula and eastern Ellsworth Land
Volcanic group Formation Distinguishing lithological characteristics Exposed thickness
Age
(Ma) Interpretation
James
Ross
Island* Hobbs Glacier** Matrix-supported conglomerate to pebbly mudstone Mainly
<
4
m
but
9.9
(one locality) Glacimarine sedimentation close to ice
(diamictite), locally fossiliferous, and laminated
tuffaceous sandstone periglacial delta-front (sandstones)
possibly up
to
25
m grounding line
(diamictitekonglomerate);
Cockbum Island Clast-supported conglomerate, pebbly sandstone
and medium
10
coarse sandstone; fossiliferous
>
10m
3-2.8
Shallow marine
(<
100
m); interglacial
(unnamed Multiple superimposed lava-hyaloclastite breccia Up
to
600
m
<=
10;
mainly
<
7
Multiple subaerial and subaqueous
volcanic strata) couplets; rare pillow lava, lapillistone; thick
lapilli tuff outcrops eruptions; englacial and marine
Seal Nunataks Bruce Nunatak Multiple and single large dykes, pillow lava,
gravelly sandstone
Christensen Nunatak Compound lava, lapillistone, lapilli-tuff;
minor agglutinate
Interbedded lava and scoria
Argo
Point
Up to
200
m
4-<
1
Entirely subaqueous; pillow volcanoes
and tuff cones (englacial)
Bellingshausen Sea Mount Pinafore Interbedded volcanic (lava, hyaloclastite breccia)
60-90
m
m
and volcaniclastic (sandstone, conglomerate) lithofacies.
Mainly red and black coarse tuff, lapillistone
and agglutinate; thin clastogenic lavas
Hornpipe Heights
20
m
0.7
(one locality) Mainly cinder cone remnants; subaerial
Upto100m
175
m
0.8-1.6
Degraded cinder cone; subaerial
7.7-5.4
Valley-confined, subglacial
‘outflow
facies’
2.7-2.5
Entirely subaerial; Strombolian
Overton Peak Buff gravelly sandstone, minor brown mudstone
I00
m
5.4
Subaqueous
tuff
cones (englacial?)
Mussorgsky Peaks Yellowish gravelly sandstone; minor mudstone,
200
m
pillow breccia, pillow
lava;
thin dykes
2.5,
0.68
(two localities) Subaqueous tuff cones (englacial)
Mount Grieg Hyaloclastite breccia, compound lava(?)
100
m undated Lava (hyaloclastite-) deltas
Venus Glacier Basaltic dykes and sills; all lamprophyres
0.8-2.3
m
15
(one locality) Minor subvolcanic intrusions
Mount Benkert Mainly gravelly sandstone; minor pillow lava,
350
rn
pillow breccia, thin massive lava undated Subaqueous tuff cones (englacial?)
Mount McCann Interbedded lava and scoria
60
m undated Degraded cinder cones
Henry Nunataks Interbedded has; minor (?)hyaloclastite breccia
12-100
m
6
(one
locality) Uncertain (too little information)
*only sedimentary formations formally defined in the James
Ross
Island Volcanic Group; volcanic strata remain to be assigned in
a
formal stratigraphy
**tentatively includes numerous poorly investigated, undated outcrops mainly
in
northern James
Ross
Island
Tt
\D
m

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY
OF
MONOGENETIC VOLCANIC FIELDS
365
At least two additional phases, now concealed beneath the ice
cap on Mount Haddington, were also postulated based
on
geomorphological evidence. However, no formations were
formally defined. Correlations were based largely on the
relative heights above sea level of each lava-breccia couplet,
an assumption of no majorfaulting, the occurrence ofdistinctive
features
in
some units (e.g. overlap structures,) that are of
genetic significance and were assumed to be stratigraphically
confined, and using
a
petrological model of olivine
compositions postulated,
on
minimal evidence,
to
evolve
progressively through the volcanic sequence (from forsteritic
compositions in the earliest phases, to more fayalitic
compositions in the latest phases). The stratigraphical model
was supported by observations particularly
on
the south-
eastern coast of James Ross Island, where up to five volcanic
phases are clearly superimposed. However, Nelson (1975)
acknowledged the difficulty of correlating over large distances,
whereexposure is incomplete, because oftheclose lithological
similarities between the individual lava-breccia couplets.
Large- and small-scale faults also affect the sequence (Nelson
1975, del Valle
&
Rinaldi 1992), and there is field evidence
that the Mount Haddington volcano has deformed the
underlying ductile Cretaceous sedimentary basement and
volcanic units in a broad anular outcrop (unpublished
information of J.L. Smellie
&
A.P.M. Vaughan). The
deformation style is unclear at present but probably includes
faulting, thrusting and local uplift similar to deformation
associated with volcanoes that have gravitationally settled
Fig.
2.
Sketch map showing the
distribution
of
volcanic
outcrops and principal
lithofacies
in
the James
Ross
Island Volcanic
Group,
northern Antarctic Peninsula
(see Fig.
1
for
location).
into their substrates (cf. Borgia 1994, van Wyk de Vries
&
Borgia 1996), and possibly block faulting related
to
regional
tectonics (del Valle
?k
Rinaldi 1992). Thus, an absence
of
stratigraphical complications caused by deformation of parts
of
the JRIVG cannot be assumed. Sykes (1988a) also
demonstrated that olivine compositions throughout the JRIVG
do not vary systematically through the sequence and, with the
large database of K-Ar isotopic ages now available, it is
evident that many of the stratigraphical correlations made by
Nelson (1 975) are untenable (Fig.
3).
For the sedimentary strata present,
two
formations have
been erected: theHobbs Glacier Formation (Pirrie
etal.
1997)
and Cockburn Island Formation (formerly the
“Pecten-
conglomerate” of Anderson (1 906); Jonkers 1998a, 1998b,
Jonkers
&
Kelley 1998). However, only the sedimentary
outcrops at Cockburn Island and south-eastern James
Ross
Island have been examined
in
detail. There are numerous
additional outcrops known, particularly
on
Ulu Peninsula,
which have only been examinated at a reconnaissance level
(Bibby 1966, Sykes 1988a, del Valle
et
al.
1987) and whose
field relations, age, correlation and lithological variations
of
the sedimentary strata are very poorly known.
Because of these problems, it is not yet possible to propose
a comprehensive lithostratigraphy for either the volcanic or
sedimentary rocks
in
the JRIVG, nor to describe fully the
lithofacies variations within any
of
the stratigraphical units
currently envisaged. For thcse reasons, no formal description
of
the constituent units
of
the
JRIVG
is attempted here,

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
366
J.L.
SMELLIE
9
Phase
5
a
Fig.
3.
Summary
of
isotopic
ages
for
lavas from the James
Ross
Island
Volcanic Group.
The
data are grouped according to the
stratigraphical “volcanic phases”
of
Nelson
(1
975).
Despite the
presence
of
at
least five superimposed volcanic phases
(e.g.
observed on the flanks
of
Mount Haddington), there is no
obvious support
for
the published stratigraphy.
All
ages
determined
by
the
K-Ar
method.
Error
bars are
20.
Includes
unpublished isotopic ages of the author.
beyond this summary of the entire group.
Thickness:
At least 1470 m, at Mount Haddington.
Lithology, petrology and palaeoenvironment:
The JRIVG
comprises pillow lava, numerous superimposed lava-
palagonitised hyaloclastite breccia couplets, several smaller
but geographically widespread outcrops of lapilli tuff, and
minor lapillistone (Nelson 1975, Sykes
1988,).
Several
small
pyroclastic cones are present on the south-east side of the
island (Strelin
&
Malagnino 1992). Pillow lava is rare and
restricted mainly to
Ulu
Peninsula, where it
is
up
to
15
m thick
and mainly occurs within and
at
the base of hyaloclastite
breccia units. The volumetrically
dominant lava-palagonitised
hyaloclastite breccia couplets are individually
20-1
80
m
thick. In places, breccia is superimposed on breccia and the
intervening lava has been eroded prior to deposition of the
youngerbreccia. The lavas are compound (pahoehoe), forming
horizontal sheets formed of up to
30
flow lenses, each lens
being
1-5
m thick. Lava lenses are locally interhedded with
the topmost few metres of the thick, steep-dipping
(25-35”)
hyaloclastite breccia beds. The breccias are homoclinally
planar stratified, or wavy stratified on
a
scale of
50-100
m.
Beds are massive and
2-5
m thick. They locally contain lenses
up to
30
m in length composed of thinly stratified fine lapilli-
tuff-grade hyaloclastite. Massive coarse polymict (lithic)
breccias up to
20
m thick and extending laterally up to
1
km
also
occur at the base of some hyaloclastite breccia units.
Sequences
of
lapilli tuff occur at several localities, with large
outcrops present at Tortoise, Terrapin, Sungold and San
Fernando hills, and Bibby Point (Fig. 2). The deposits exceed
600 m in thickness and comprise thinly and crudely stratified
monomict lapilli tuff showing
a
variety of sedimentary
structures, including loading and flame structures, low-angle
cross stratification, ripples, scours and deformed (slumpccl)
beds. Monomict lapillistones have
a
restricted distribution,
occurring
as
thin planar beddcd red-coloured units within
lava
sequences
at
a
few localities.
Sedimentary strata, described
as
“tuffaceous conglomerates“,
“marine tuffs” and “diamictites” by Bibby (1966), Nelson
(1975) and Sykcs
(1988a)
and tentatively assigned
to
thc
Hobbs Glacier Formation by Pirrie
et
al.
(1
997), have
bccn
described
at
several localities
on
James Ross Island.
Thc
deposits are polymict, with
a
mixed Antarctic Peninsula ant1
James Ross Island provenance, and comprise poorly sortcd
matrix-supported conglomerate to pebbly mudstone diamictitc
with very poorly preserved marine fossils, and laminatcd
volcanic sandstone with only
a
James
Ross
Islandprovcnancc
(described by Pirrie
et
al.
(1997)
as
sand-
to
silt-grade
tufl).
The sequences are typically just
a
few metres thick, possibly
ranging up to
25
m
in
northern James Ross Island. By contrast.
the Cockburn Island Formation consists of more than
I0
ni
of’
bedded medium to coarse sandstone, pebbly sandstone
and
pebble to boulder conglomerate almost exclusively derivctl
from
a
JRIVG provenance and containing numerous niarinc
fossils, some reworked probably from Eocene and Crctaceous
strata now exposed on Seymour Island to the east (Jonke~-:;
1988a).
A
sequence composed of polymictconglomeratc
3
11-1
in thickness, broadly similar in lithology and structural position
to the Cockburn Island Formation
but
with
a
mixed Antarclic.
Peninsula and James Ross Island provenance like the
Hobha
Glacier Formation,
also
crops out at
c.
60
m
above sea
lcvci
on the north-east coast of James Ross Island (fiord0 BelCii.
del Valle
etal.
1987). Itsstratigraphical affinities are uncertaiii
(Jonkers 1998b).
The cogenetic lava-hyaloclastite brecciacouplets (volcanic
phases of Nelson
1975)
represent multiple superimposed
hyaloclastite deltas, formed when subaerial lava(s) entered
ponded water, whereas the lapilli tuff outcrops represent
tul‘l’
cone successions erupted and deposited cither subaerially or
subaqueously, possibly within the sea or
in
englacial vaults
(cf. Nelson 1975, Pirrie
&
Sykes 1987, Sykes
1988a,
Smcllic
etal.
1988, Skilling 1994, Smellie
&
Skilling 1994). Some
ol
the structural relationships observed
in
the volcanic sequences
(e.g. various “ovcrlap structures” of Nelson 1975) indicate
fluctuating water levels cocval with individual eruptions.
They are hard to explain in terms of
a
marine setting
and
changes
in
sea level (cf. Nelson 1975) and are probably
diagnostic of englacial eruptions (cf. Smellie
in
press
a,
in
press b). However, the presence of
in situ
marine fossils
in
some sedimentary beds and the compositions
of
authigcnic
phases in the hyaloclastite breccias (palagonite and phillipsitc;
unpublished information of British Antarctic Survey) suggest
that somepartsof the sequence weremarine. Diamictite
in
thc
Hobbs Glacier Formation was
also
interpreted
as
a
product
of
glacimarine sedimentation close to
a
glacier grounding
linc
(Pirrie
et
al.
1997). Thus, it
is
likely that the eruptive
and
depositional environment varied between marine and englacial
during the
c.
IOmillion yearpcriod represented by the JRIVG.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY OF MONOGENETIC VOLCANIC FIELDS 367
The JRIVG contains the widest compositional range of any
alkaline volcanic outcrop in the Antarctic Peninsula region.
The lavas are mainly
hy-
and ne-normative tholeiites, alkali
basalts and hawaiites, together with minor basanite and
mugearite (Nelson 1975, Smellie 1987, Sykes 1988a, Hole
etal. 1995). The most evolved compositions are in pegmatites
found in thick sills, which range
up
to tephriphonolite.
Boundaries:
An angular, erosional basal unconformity with
Cretaceous sedimentary strata
(Campanian-Maastrichtian;
Crame
et
al.
1991) is exposed around the margins of James
Ross Island, on Vega and Cockburn islands (Bibby 1966,
Nelson 1975). The unconformity displays considerable pre-
volcanic relief (rising to 213 m
in
western James Ross Island
and up to 330 m on Vega Island, Bibby 1966, Sykes 1988a).
In Graham Land, the base of the sequence is unexposed but it
is inferred to be an angular unconformity cut
in
deformed
metasedimentary strata of the Permo-Triassic Trinity Peninsula
Group (Aitkenhead 1975, Nelson 1975). The upper boundary
of the JRIVG is erosional and corresponds to the present-day
surface.
Within the JRIVG, the Hobbs Glacier Formation intervenes
locally between the basement unconformity and the
conformable base of the first overlying volcanic
unit
of the
JRIVG. It is regarded as the oldest lithostratigraphical division
of the JRIVG (Pirrie et
al.
1997). By contrast, the Cockburn
Island Formation (confined to Cockburn Island)
unconformably overlies JRIVG lava and tuff and its upper
surface
is
eroded (Jonkers 1998a, 1998b, Jonkers
&
Kelley
1998). The fiordo Beltn conglomerates occur within the
JRrVG volcanic sequence but the precise stratigraphical context
is uncertain (del Valle
et
al.
1987, Jonkers 1998b).
Age:
There are numerous isotopic ages for the JRIVG, almost
all obtained by the K-Ar method (summarized in Fig.
3,
Rex
1976, Massabie
&
Morelli 1977, Malagnino
et
al.
1978,
Smellie
et
al.
1988, Sykes 1988b, Lawver et
al.
1995). The
oldest
in
situ volcanic rocks are hypabyssal intrusions (a plug
andadyke), datedat 6.45 40.60and6.8 *0.5Ma, respectively.
K-Ar ages obtained on volcanic clasts
in
associated
sedimentary rocks extend the age of volcanism to 7.13
*
0.49
Ma. The pyroclastic cones on the south-eastern side of the
island are largely unmodified by erosion and may be very
young (post glacial?, Strelin &Malagnino 1992). By contrast,
strontium isotopic ages on shelly material
in
the Hobbs
Glacier Formation suggest that
it
may range back
to
9.9
k
0.97
Ma(Dingle &Lavelle 1998) and the presence of olivine basalt
clasts in the diamictites and interbedded tuffaceous sandstones
suggests that the initiation of eruptions in the JRIVG may pre-
date 10 Ma. By contrast, Ar-Ar ages and a fossil pectinid
fauna suggest that the Cockburn Island Formation is only 2.8-
3 Ma (Jonkers
&
Kelley 1988). The conglomerate sequence
at fiordo BelCn has yielded a Sr isotopic age on pectinid fossil
material of 6.8
+
1.3L0.5 Ma (M. Lavelle,
in
Jonkers 1998b).
Seal
Nunataks
Volcanic
Group
Name
and
history:
This is a new lithostratigraphical
unit,
named after the principal outcrop area for the volcanic
formations. Seal Nunataks were visited first
in
1893, when
eruptions on two nunataks wererecorded, although the account
is open to ambiguous interpretation (Smellie 1990a). Published
details of the lithofacies and local stratigraphies are sparse.
Distribution:
Seal Nunataks; Argo Point, Jason Peninsula
(Fig. 4).
Subdivision:
The group is divided into the Bruce Nunatak,
Christensen Nunatakand Argo Point formations. Theirrelative
ages overlap and they were coeval
in
part. The ages and field
relationships suggest that the Christensen Formation overlies
and is generally younger than outcrops of the Bruce Nunatak
Formation.
Thickness:
A maximum thickness of about 650 m is possible
assuming continuity of outcrop down
to
bedrock beneath the
Larsen Ice Shelf.
Lithology, petrology and palaeoenvironment:
The group is
entirely volcanic-derived. It is formed mainly of pillow lavas
and dykes together with smaller outcrops of subaqueously-
deposited hydrovolcanic tephra, subaerial lavas and
Strombolian scoria. The lavas are
Q-,
hy-
and slightly
ne-
Fig.
4.
Sketch
maps
showing
the
distribution
of
volcanic
outcrops
of
the
Seal
Nunataks
Volcanic
Group
at
a.
Seal
Nunataks
and
b.
Argo
Point.
The
volcanic
outcrops
are
coloured solid
black
(also
indicated
by
arrow
in
b).
Exposures
of
older
rock
formations
are
indicated
without
ornament. See Fig. 1
for
the
locations
of
the
outcrops
in
the
Antarctic
Peninsula.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
368
J.L.
SMELLIE
normative alkali basalts, hawaiites and tholeiites (Saunders
1982, Hole 1990a). The Christensen Nunatak and Argo Point
formations are entirely subaerial, whereas the Bruce Nunatak
Formation is subaqueous. An englacial palaeoenvironment is
likely for the Seal Nunataks outcrops (Smellie
&
Hole 1997).
Boundaries:
The lower boundary is unexposed but an
unconformable contact, probably with Jurassic-Cretaceous
back-arc basin sedimentary rocks (Seal Nunataks) or Jurassic
volcanic rocks (Jason Peninsula), can be inferred from field
relationships and the presence of rare metasedimentary
xenoliths in the volcanic rocks (Fleet 1968, del Valle
et al.
1983, 1997, del Valle
&
Medina 1985, Smellie 1991, Riley
et al.
1997). The upper boundary is erosional (present-day
surface).
Age:
4-<I
Ma(by K-Arisotopic dating, del Valle
etal.
1983,
Smellie
et al.
1988). Ages are apparently younger in the more
central nunataks and an age progression was inferred by
Gonzilez-Ferrin (1983).
Bruce Nunatak Formation
Name and history:
After Bruce Nunatak (65"05'S,
60"
14'W),
one of the larger nunataks situated on the south-western side
of Seal Nunataks.
Distribution:
Akerlundh, Arctowski, Bruce, (?)Castor (lower),
Evensen, Gray, Larsen, Murdoch, and Oceana nunataks;
Lindenberg Island (Fig.
4).
Principal lithological characteristics:
Most outcrops consist
of multiple or single large dykes (up to 70 m wide) or thick
piles of pillow lavalacking intergillow sediment. Afew dykes
are pillowed and the volcanic succession was probably dyke-
fed and erupted from fissures. Some pillow lava outcrops (at
Bruce, Gray, Murdoch and Oceana nunataks) are overlain
depositionally by buff-coloured gravelly sandstone and
sandstone-mudstone strata, which commonly show syn-
sedimentary faults and slump structures. Most of the strata
correspond to the pillow volcano stage of subglacial and
subaqueous volcanoes.
Thickness:
About
150
m
of rock is exposed in Murdoch
Nunatak, but a maximum thickness of about 650 m is possible
for the formation as a whole.
Boundaries:
The lower boundary is obscured by the Larsen
Ice Shelf but the surface is thought to be an unconformity with
Upper Cretaceous back-arc basin sedimentary rocks. The
upper boundary corresponds to the present-day erosion surface,
or is unconfomably overlain by subaerial volcanic rocks of
the Christensen Nunatak Formation.
Age:
4-<0.2
Ma (K-Ar). The ages cluster mainly around
1
.S
Ma and the oldest age
(4
k
1
Ma) may be unreliable. Not
all
of
the outcrops have been dated.
Type section:
Bruce Nunatak consists of three ridges formed
around multiple dyke zones trending NNE, NE and
ESE.
Pillow lavas predominate on the flanks and lower slopes of the
ridges. The upper parts of the northern ridge consist
of
yellow-weathering gravelly sandstone with lenses of pillo~
breccia and rare flattened glassy bombs. The well-developed
planar stratification is severely disrupted by syndepositional
faulting and major, large-scale, northerly-directed slumping
Bedding involved in the latter shows dips up to
54"
and parts,
of the sequence are partly disaggregated, comprising
disorientated more resistant blocks encased in massive sand)
matrix.
Key references:
Fleet (1968), Gonzalez-Ferrin (1983),
dcl
Valle
et al.
(1983), Smellie
et
al.
(1988), Hole
(1990a),
Smellie (1990a), Smellie
&
Hole (I 997).
Christensen Nunatak Formation
Name:
After Christensen Nunatak (65"06S, 59"3 l'W), situated
at the easternmost limit of Seal Nunataks.
Distribution:
Mainly Bull, Castor (upper), Christensen,
Dallmann, Donald and Hertha nunataks; also parts
of
Akerlundh, Arctowski, Evensen, Murdoch and Oceana
nunataks and Lindenberg Island (Fig.
4).
Principal lithological characteristics:
Mainly subaerial
ly
erupted compound lavas and buff or grey lapillistones and
lapilli tuffs. The lapilli tuffs are hydrovolcanic deposits.
Thcy
are
>80
m thick at Castor Nunatak and form a crater-like
structure with a surface depression 60 m deep and 150
m
wide
infilled by horizontal lavas. Lapillistones at Donald Nunatak
arecinderconedeposits unusually altered to yellow palagonitc
along subvertical joints. Also present are isolated
small
outcrops of black weakly welded agglutinate ('spatter'). Thet-c
is
noconvincing evidence forthe primary landforms,
fumarole.;
and recent eruptions reported. The outcrops are believed
to
represent the subaerial caps to volcanoes erupted beneath
thick ice sheets.
Thickness:
The largest outcrop, at Christensen Nunatak,
shows about
100
m of exposure
but
much
is
obscured by scrcc.
A possible thickness of 200m
of
likely compound lavas crops
out at Bull Nunatak. The total thickness of the formation
is
unknown but is probably much less
than
the Bruce Nunatak
Formation.
Boundaries:
The lower boundary is obscured by the Larscn
Ice
Shelf or is an unconformity on the Bruce Nunatak
Formation. Isolated outliers of agglutinate and a possiblc
degraded spatter cone rampart overlie the Bruce Nunata
k
Formation at Murdoch Nunatak. Thc upper boundary is an
erosional surface.
Age:
Only dated at Christensen Nunatak (0.7
2
0.3
Ma
by
K-Ar).
Type
section:
Christensen Nunatak, rising to 305 m abovc
sca

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY OF MONOGENETIC VOLCANIC FIELDS 369
level, consists of two subhorizontal lava flows interbedded
and overlain by yellowish-coloured lapilli tuffs. Evidence for
lava-hyaloclastite sequences, similar to those well developed
on James Ross Island, has not been confirmed by more recent
work. The
1owerlavaatChristensenNunatakis
20m thickand
columnarjointed, whereas the upper lavais only 3 m thick and
massive. Both lavas are highly vesicular and display brick-
red, ropy scoriaceous upper surfaces. The interbedded
volcaniclastic rocks are mainly subhorizontal, planar-stratified
lapilli tuffs
of
likely hydrovolcanic (tuff cone) type. Bomb
sags are developed around some large accessory blocks. The
tuffs reach a maximum thickness of about
50
m
between the
two lavas. There are also minor dykes with no preferred
orientation.
Key references:
Fleet (1 968), Gonzhlez-FerrBn (1983), del
Valle
et al.
(1983), Smellie
et al.
(1988), Hole (1990a),
Smellie (1990a), Smellie
&
Hole (1997).
Argo Point Formation
Name:
After Argo Point (66" 15'S, 60"5S'W), asmall headland
rising
c.
260 m above the Larsen Ice Shelf, on the south-east
side of Jason Peninsula.
Distribution:
Argo Point (Fig.
4).
Principal lithological characteristics:
Argo Point consists
of
a degraded cinder cone about 300 m in diameter overlying
inaccessible cliffs of interbedded lava and scoria of similar
age to the cinder cone.
Fig.
5.
Sketch maps showing the distribution
of
known (solid black) and inferred (grey)
volcanic outcrops of the Bellingshausen Sea
Volcanic Group in Alexander Island.
a.
Rothschild Island,
b.
Mount Pinafore,
Debussy Heights and Hornpipe Heights,
c.
Beethoven Peninsula,
d.
South-eastern
Alexander Island (star symbols indicate
dyke andlor sill localities). Small outcrops
in a and
b
are
also indicated by
arrows.
Exposures of older rock formations are
indicated without ornament. See Fig.
1
for
the locations
of
the outcrops within
Alexander Island.
Thickness:
>
c.
175 m.
Boundaries:
The Argo Point Formation is probably
unconformable on Jurassic volcanic rocks but the base is
obscured by ice. The upper boundary
is
an erosional surface.
Age:
Two basalts from Argo Point yielded K-Ar whole-rock
ages of 0.8-1.6 Ma.
Type section:
Argo Point.
Key references:
Saunders (1982), Smellie
et al.
(1988),
Thomson
(1
990).
Bellingshausen Sea Volcanic Group
Name and history:
Outcrops included in this group are scattered
widely across Alexander Island and also occur at Snow
Nunataks, Sims Island, Rydberg Peninsula and Merrick
Mountains (Figs
5
&
6).
All
were included
in
the
"Bellingshausen Sea Volcanic Province" by LeMasurier
&
Thomson (1990) and the name is retained here, modified
to
conform with lithostratigraphical nomenclature. The outcrops
are of recent discovery and were described by Horne
&
Thomson (1967), Bell (1973), Care (1980), Burn
&
Thomson
(1981), O'Neill
&
Thomson (198S), Thomson
&
Kellogg
(1990) and Rowley
et al.
(1990). Detailed studies
of
the
geochemistry and physical volcanology of the Alexander
Islandoutcrops were published by Hole( 1988,1990b), Smellie
et al.
(1993), Smellie
&
Skilling (1994) and Smellie
&
Hole
(1997). The Mount Pinafore Formation outcrops have been
proposed as sequence holotypes for products of subglacial

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
370 J.L. SMELLIE
eruptions of 'sheetflow type', erupted beneath valley glaciers
(Smellie et al. 1993).
Distribution: North, south-west and south-east Alexander
Island; south-east Rothschild Island; Beethoven Peninsula;
south-east Alexander Island; Snow Nunataks, probably Sims
Island and Rydberg Peninsula; Merrick Mountains.
Subdivision: Nine formations are defined: Mount Pinafore,
Hornpipe Heights, Overton Peak, Mussorgsky Peaks, Mount
Grieg, Venus Glacier lamprophyres, Mount Benkert, Mount
McCann and Henry Nunataks formations. The Venus Glacier
lamprophyres consist entirely of dykes and sills.
Thickness: The thinnest sequence occurs on the Merrick
Mountains (<I2 m) and the thickest at Beethoven Peninsula
and Snow Nunataks (at least 300400 m, possibly much
greater than
400
m assuming continuity of outcrop down to
pre-volcanic bedrock; Smellie et
al.
1988, 1993, Smellie
&
Hole 1997). Intrusions of the Venus Glacier Formation are
generally <1
.O
m
in
thickness (Horne
&
Thomson 1967 and
unpublished field notes of P.A. Doubleday, British Antarctic
Survey).
Lithology, petrology and palaeoenvironment: The group is
entirely volcanic, formed of dykeskills, basaltic lavas and
a
variety of subaerial and subaqueous volcaniclastic lithofacies.
Eruptive palaeoenvironments ranged from subglacial (beneath
thin valley glaciers at Mount Pinafore/Debussy Heights;
beneath
a
thickice sheet at Beethoven Peninsula and probably
Overton Peak (Rothschild Island) and Snow Nunataks) to
entirely subaerial (Hornpipe Heights, Hole 199Oc, Smellie
et al. 1993, Smellie
&
Hole 1997). The volcanic outcrops
form two contrasting compositional groups: hy- and slightly
ne-normative alkali basalts, hawaiites and tholeiites (Beethoven
Peninsula; Snow NunataksIRydberg Peninsula) and strongly
undersaturated ne-normative basanites and phonotephrites
(all
other localities, Srnellie 1987, Hole 1988, 1990b, 1990c,
Hole
&
Thomson 1990, Rowley
&
Thomson 1990, Thomson
&
O'Neill 1990, Hole et
al.
1991b). The Venus Glacier
lamprophyres comprise compositionally distinctive alkali
basalts, potassic trachybasalts and phonotephrites (Rowley
&
Smellie
1990andunpublisheddataoftheauthor),
andthey are
also
petrographically different from other Cenozoic alkaline
rocks in the Antarctic Peninsula (lamprophyres, Horne
&
Thomson 1967).
Boundaries: Most of the volcanic outcrops on Alexander
Island unconformably overlie deformed metasedimentary
rocks
of the LeMay Group (Jurassic-Cretaceous, Holdsworth
&
Nell 1992), although contacts at Beethoven Peninsula and
Rothschild Island are obscured by ice. An outcrop
at
thc
summit of Mount Pinafore unconformably overlies
calc-
alkaline arc volcanic sequences of the subduction-relakd
Alexander Island Volcanic Group (Elgar Formation, Carl!/-
middle Eocene, McCarron
&
Millar 1997). Dykes
in
thc
Vcnus Glacier area intrude middle Cretaceous formations
of
the Fossil Bluff Group (Horne
&
Thomson 1967, Moncrieff&
Kelly 1993). In each outcrop area, the alkaline volcanic
and
dyke formations are the youngest rocks present and their
upper boundary corresponds to the present erosion
surfacc.
Age: Three groups of ages (K-Ar) areevident, corresponding
to 15 Ma (Venus Glacier), 7.7-5.4 Ma (Mount Pinalod
Debussy Heights, Rothschild Island, Merrick Mountains) and
2.7-<1 Ma (Beethoven Peninsula, Hornpipe Heights, Smcltic
et al. 1988). Not all of the outcrops have been dated.
Mount Pinafore Formation
Name: AfterMountPinafore (69"46'S, 70°58'W), apromincnt
mountain massif rising to
c.
1100
m in northern Alexander
Island.
Distribution: Ravel Peak (Debussy Heights) and threelocalities
on
Mount Pinafore (Fig.
5).
Principal lithological characteristics: Multiple sequcnccs
mainly composed of two principal lithofacies associations,
a
basal association of volcaniclastic lithofacies (sandstone,
conglomerate) and an upper association of volcanic lithofacies
(lava, hyaloclastite breccia). The two associations
form
cogenetic "couplets" that are superimposed up to four
times
locally (e.g. Debussy Heights). The volcaniclastic sections
comprise beds of hydroclastic tephradeposited predoniinanr
I
y
from traction currents and
as
mass flows. The reworking
was
coeval with eruptions. Many beds
also
contain
a
variety
of
clasts derived from the underlying pre-volcanic bedrock and
associated diamictites were possibly derived from
cocval
glaciers. The lava-hyaloclastite breccia lithofacies represent
basalt effusion and generation of breccia by chiIling
(thennal
shock), granulation and spallation during quenching by contact
with glacier ice and meltwater, probably in valley-confined
Fig.
6.
Sketch maps showing
the
distribution
of
volcanic outcrops
of
the
Bellingshausen Sea Volcanic
Grou"
in
eastern Ellsworth Land.
a.
Snow
Nunataks, Rydberg Peninsula
and
Sims
Island,
b.
Merrick Mountains.
Symbols
and
ornaments
as
in
Fig.
3.
See Fig.
I
for
locations
of
the
outcrops.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY
OF
MONOGENETIC VOLCANIC
FIELDS
37
1
subglacial tunnels. The deposits are crudely comparable to
eskers.
Thickness:
All the sequences are relatively thin, typically just
a
few tens of metres, ranging locally to
c.
70 m.
Boundaries:
Basal contacts are usually well exposed. The
underlying pre-volcanic surfaces are typically steep-dipping
and show polishing, striations and molding owing to erosion
by Miocene valley glaciers coeval with the alkaline volcanism.
The upper surface
is
erosional.
Age:
K-Ar
whole-rock isotopic dating of the three sequences
at Mount Pinafore and that at Ravel Peak indicate eruptions in
latest Miocene times (7.7-5.4 Ma). A significantly younger
age of 3.9 Ma obtained from one of the Mount Pinafore
outcrops
is
likely to be unreliable (reset).
Type section:
The outcrop on the south-western spur ofMount
Pinafore is the type section for this formation. It is described
in
detail by Smellie
&
Skilling (1994).
Key references:
Bum &Thornson (198
l),
Hole (1988), Smellie
et al.
(1988,
1993),
Hole
&
Thomson (1990), Hole
et al.
(1991b), Smellie
&
Skilling (1994).
Hornpipe Heights Formation
Name:
After Hornpipe Heights (69"52'S, 70"3S'W), an
irregular ridged massif rising to
c.
1200 m south of Mount
Pinafore, northern Alexander Island.
Distribution:
Only
a
single outcrop is included in this
formation. It is present on the north-western side of Hornpipe
Heights overlooking Sullivan Glacier (Fig.
5).
Principal lithological characteristics:
There is little published
information on the sequence
at
Hornpipe Heights and the
opportunity is taken here to provide a more detailed description
(based on unpublished field notes
of
M.J. Hole. British
Antarctic Survey). The sequence has
a
general dip of
c.
40"N
(range typically 35-45") except locally where it drapes
upstanding knolls in the underlying bedrock. The (?)lowest
bed (upper contact erosional) is
a
polymict orthobreccia
composed of abundant angular blocks of LeMay Group
sandstone and basalt, vesicular lapilli and
a
sandy matrix. It
has
a
crude statification that becomes better-defined upwards
by concentrations of coarser and finer clasts. The next bed up-
sequence is
a
widespread black lapillistone, which is up to3 m
thick in hollows but thins up-dip and over basement highs
to
<0.5
m.
It
is
overlain by
a
varied sequence about 7 m thick
composed of alternating yellowish, buff- and red-coloured
coarse tuffs and lapillistones, the latter with dispersed bombs
commonly showing flattening parallel to bedding, crude planar
stratification and reverse grading. Normal grading is
also
rarely present and many beds are truncated by minor
unconformities. Some finer beds
also
show delayed reverse
grading confined to the top parts of beds and rare finer tuffs
show ripple-like small-scale folds that verge up-dip. Steep
faults are common and they may be draped by lapillistone
beds. Higher parts of the sequence, up to
c.
20
m thick,
comprise thick beds of striking red agglomerate (agglutinate)
mainly formed of large ovoid ropy-textured bombs with
fluidal aerodynamically molded surfaces and minor coarse
lapilli. The beds commonly coarsen and thicken upwards and
they are weakly welded. The agglutinate sequence is overlain
along uneven channel-like surfaces up to
50
m wide and
10
m
deep by multiple thin (generally
1-2
m thick) lavas, many of
which alternate with thin (0.3-0.5 m) agglutinates. The lavas
thin up-slope. They rarely show poor columnar jointing but
platy bed-parallel jointing
is
ubiquitous. A yellow discoloration
is evident in the topmost few dm of lapillistones underlying
some lavas.
The scarcity of fine tuff, either
as
beds or matrix, and the
abundance of red or black highly vesicular ash and lapilli
(many with fluidal "droplet" surfaces) and bombs are
characteristics of fall tephra, the products of "dry" eruptions
of low energy Strombolian type. The basal polymict
orthobreccia, with its abundant LeMay Group clasts, may
represent
a
product of initial vent-clearing eruptions; the
proportion of basement-derived accidental clasts diminishes
rapidly above the basal bed but such clasts are seldom
completely absent. The common reverse grading, thinning
over basement highs and multiple shallow unconformities
were probably caused by local remobilization
(as
grain flows)
of unstable beds of cohesionless sand- and gravel-grade
pyroclasts shortly after deposition on the very steep substrate
(the angle of dip exceeds the stable angle
of
repose for
cohesionless sediments). Other evidence for contemporaneous
slope instabilty consists of syndepositional faulting, minor
slumping (ripple-like folds?) and channelized surfaces. The
lavas are probably clastogenic flows whose ubiquitous platy
appearance is probably caused by shearing during laminar
flow of
a
viscous liquid. The yellow discoloration of some
beds is caused by marginal palagonitization of sideromelane
and tachylite. Such alteration is most commonly found in
hydrovolcanic and subaqueous tephra, but there is no evidence
for hydrovolcanism or subaqueous deposition.
The
alteration
is
marginal to clasts and may be due to surface weathering
in
amoist environment. However, the conspicuous restriction of
some alteration to the upper parts of beds overlain by lavas
also
suggests the possibility of
a
local snow cover melted by
the lavas, the small quantities of meltwater released then
causing the alteration. The location of the vent responsible
is
unknown but the overall coarseness of the sequence, abundant
large bombs and clastogenic lavas suggest that it was within
a
few hundred metres. Although the presence of lavas on the
topographically higher parts of the outcrop could suggest
a
vent close to the ridge crest,
a
vent locus beneath Sullivan
Glacier seems more likely.
Thickness:
Up to
20
m of the formation is preserved.
Boundaries:
The lower boundary is
a
highly uneven, steeply

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
372 J.L. SMELLIE
dipping (c. 45") erosional surface formed in deformed
metasedimentary rocks of the LeMay Group; several
upstanding rocky knolls of LeMay Group are preserved and
are draped by the Hornpipe Heights Formation. The surface
shows no signs of pre-volcanic glacial erosion. The upper
boundary conforms to the present day erosion surface. The
outcrop
is
obscured down-dip by Sullivan Glacier.
Age: Two basanite lavas have yielded K-Ar isotopic whole-
rock ages of 2.5
?
0.8
and 2.7
?
0.2 Ma.
Type section: Outcrop at Hornpipe Heights, described above.
Key References: Smellie et al. (1988), Hole (1990~).
Overton Peak Formation
Name: After Overton Peak (69"41'S, 71"58'W), one of the
DeskoMountains rising
toc.
550m on south-eastern Rothschild
Island.
Distribution: Two rounded, predominantly scree-covered
hills and
a
100
m-high cliff situated north-west and west of
Overton Peak, respectively (Fig.
5).
Principal lithological characteristics: Mainly crudely bedded
and normal-graded yellowish gravelly sandstones and minor
brown mudstones showing faintplanar stratification and cross
lamination, respectively, in beds 0.2-1 m thick. Large- and
small-scale faults are conspicuous and wash-out channels are
locally common. Bedding dips generally at 18-25' but
orientations are very variable suggestive of post-depositional
slumping. Three basalt dykes are
also
present, one of which
is c.
50
m wide. Although studied only at reconnaissance
level, the Overton PeakFormation appears to closely resemble
the better-described Mussorgsky Peaks Formation (below)
and a similar origin (eruptions beneath
a
thick
ice
sheet) is
inferred.
Thickness: >lo0 m.
Boundaries:
No
contacts with other rock formations arc
exposed but the lower boundary is presumed to be
unconformable on the LeMay Group and the upper boundary
is the present-day erosion surface.
Age: One of the largely scree-covered hill outcrops has
yielded an age of 5.4
2
0.7 Ma (K-Ar).
Type section: The cliff exposure
5
km west of Overton Peak
is selected as the type section for this formation and is
described and illustrated by Care (1980).
Key References: Care (1980), Smellie et al.
(1
988).
Mussorgsky Peaks Formation
Name: After Mussorgsky Peaks (71"44'S, 73"41'W), two
prominent nunataks near the head of Brahms Inlet, Beethoven
Peninsula, south-western Alexander Island.
Distribution: Mussorgsky Peaks, Mount Liszt, Mount Grieg,
Mount Strauss and Gluck Peak. Four additional outcrops,
at
MountTchaikovsky, Mount Lee, Mount Schumann andchopin
Hill, have not been visited but may also be formed of volcanic
rocks of this formation (Fig.
5).
Principal lithological Characteristics: Outcrops are dominatcd
by crudely bedded yellowish gravelly sandstones. Othcr
lithofacies include thin-bedded sandstone-mudstone, sandy
pillow-fragment breccia and minor pillow lava; several
outcrops also contain thin basalt dykes. Large- and
smaYI-
scale syn-sedimentary faulting and slumping are common and
conspicuous. The volcaniclastic lithofacies are sedimentcd
hydroclastic tephra, representing
a
variety of mass
flows
depositedmainly from proximal high- andlow-density turbidity
currents anddebris flows. The different styles of sedimentation
present have been related
to
a model of variable eruption
dynamics during subaqueous (en glacial) tuff cone construction
in
a
subglacial setting.
Thickness: 200 m
of
section is exposed on the western
of
the
two nunataks at Mussorgsky Peaks but total thicknesses
>500
m are possible for the Beethoven Peninsula outcrops,
assuming continuity of outcrop down to sub-ice pre-volcanic
bedrock.
Bounduries: The lower boundary is unexposed (obscured
by
ice) but the formation is presumed to be unconformahlc on
LeMay Group. The upper boundary generally conforms
to
thc
present day erosion surface but, at Mussorgsky Peaks and
Mount Grieg, the formation is overlain by the Mount Grieg
Formation across an uneven undulating gently dipping
sur1iv:c.
The upper surface may be
a
syn-depositional compound
unconformity (slump scar) formed by multiple large-scale
sector collapses of the volcanic edifices.
Age: A sample from Mussorgsky Peaks was dated
as
2.5
Ma,
and another at Gluck Peak
as
0.68
?
0.97 Ma (by
K-Ar;
the
latter unpublished data of M.J. Hole). Other outcrops on
Beethoven Peninsula are undated.
Type section: The western (larger) of the two nunataks at
Mussorgsky Peaks. The section is up to 200 m high and
3
km
long and is described in detail by Smellie
&
Hole (1997).
Key references: Bell (1973), Hole (1990b), Smellie
&
Holc
(1
997).
Mount Grieg Formation
Name: After Mount Grieg (7 1"36'S, 73"l I'W), aprorninrmt
nunatak rising to c.
600
m at the head
of
Brahms Inlct.
Although the entire sequence is inaccessible, it is very
well
exposed on its north-western face where
a
section
300-400
ni
high is easily observed by binoculars; only the upper
100-
150
m corresponds to the Mount Grieg Formation.
Distribution: Western nunatak of Mussorgsky Peaks; Mount

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY OF MONOGENETIC VOLCANIC FIELDS 373
Grieg (Fig.
5)
Principal lithological characteristics:
At Mussorgsky Peaks,
the sequence forms the summit of
a
mesa. It is composed
of
dark grey glassy breccia (hyaloclastite) with abundant intact
and fragmented basalt pillows. It has awedge-shaped geometry
and crude large-scale, homoclinal, steep-dipping planar
stratification interpreted
as
foreset bedding in
a
hyaloclastite
(or lava) delta which prograded into an englacial lake. The
expected overlying subaerial lavas have been eroded away at
that locality but they are still preserved as an inaccessible
subhorizontally bedded sequence of compound(?) lavas
50-100m thick above breccias at the summit of Mount Grieg.
The contact between the lavas and hyaloclastite breccia at
Mount Grieg is hard to distinguish, but there is
a
prominent
contact between the dark grey Mount Grieg Formation
lithofacies and bright yellow deposits of the underlying
Beethoven Peninsula Formation, which extends down to the
level of the surrounding ice sheet
(a
vertical interval
of
about
200-300
m).
Thickness:
At least
100
m thick at Mussorgsky Peaks. There
is an uncertain but probably comparable thickness on Mount
Grieg.
Boundaries:
At both localities, the basal surface is uneven,
undulating and gently dipping on the Beethoven Peninsula
Formation (see description above). The upper boundary is the
present-day erosion surface.
Age:
No parts of the Mount Grieg Formation have been dated
directly but,
as
it is coeval with the Beethoven Peninsula
Formation at Mussorgsky Peaks (dated by K-Ar
as
2.5 Ma),
a
similar age is likely at that locality, at least.
Type section:
Mount Grieg (see description above).
Key
references:
Hole
(1990b),
Smellie
&
Hole (1997).
Venus Glacier Eampruphyres
Name:
After Venus Glacier (71"36'S, 68'27'W) in south-
eastern Alexander Island. Dykes/sills are very uncommon
in
eastern Alexander Island. Most occur in the vicinity of Venus
Glacier and only those are distinguished here
as
the Venus
Glacier Lamprophyres.
Distribution:
In addition to outcrops described from Waitabit
Cliffs, Cannonball Cliffs and Triton Point, alkaline dykes and
a
sill are now also known to crop out at the Quadrangle, upper
reaches
of
Venus Glacier and Horrocks Block (unpublished
field notes of P.A. Doubleday, British Antarctic Survey;
Fig.
5).
Principal litholugicaf characteristics:
Entirely formed of
basaltic dykes and sills, although they may have been feeders
for volcanic centres that have been completely removed by
erosion. They are petrographically and compositionally
distinctive lamprophyres (camptonites) with alkali basalt/
basanite-phonotephrite compositions, formed of euhedral
phenocrysts
of
titanaugite, hornblende, altered olivine and
minor plagioclase in
a
coarse groundmass of titanaugite,
hornblende, opaque oxide, olivine and pervasive chlorite,
zeolite (analcite?) and carbonate (including unpublished
informationoftheauthor). Phenocrysts ofbiotiteand kaersutite
are also present andaremuch larger than the otherphenocrysts.
Thedykes trend mainly north-eastwards
(040-078"N),
with a
subsidiary trend towards 104"N. An intrusion exposed in the
upper Venus Glacier areashows both sill anddykerelationships
(unpublished field notes
of
P.A. Doubleday, British Antarctic
Survey).
Thickness:
The intrusions are mainly
2
0.8
m thick but range
up to 2.3 m; the 7.0 m thickness reported by Horne
&
Thomson
(1 967) for one dyke was
a
printing en-or and it should be 0.7 m
(M.R.A. Thomson, personal communication).
Boundaries:
The dykes/sills intrude sedimentary rocks of the
Pluto Glacier and Neptune Glacier formations (middle
Cretaceous) of the Fossil Bluff Group. Although they typically
follow faults and fractures in the Fossil Bluff Group, at least
one dyke (at Waitabit Cliffs)
is
cut by
a
brittle shear zone.
Age:
Adykeat WaitabitCliffs yieldeda K-Arwhole-rockage
of
15?
1
Ma.
Type section:
All the intrusions show similar features but the
southern dyke at Waitabit Cliffs is probably one of the best
exposed, forming a preferentially exhumed, conspicuous wall-
like outcrop, and it is selected
as
the type example for these
intrusions.
Key
references:
Horne
&
Thomson (1967), Rex
(1970),
Rowley
&
Smellie
(1990).
Mount
Benkert Formation
Name:
A€ter Mount Benkert (73"38'S, 76"40'W), Snow
Nunataks, which rises
to
c.
700
m and
is
situated south-east
of
Carroll Inlet, English Coast (Fig.
6).
Distribution:
Snow Nunataks (Mount Benkert, Mount
Thornton, basal sequence at Mount McCann). The
geographical proximity (only 25 km distant) and apparent
lithological similarities (judged from photographs and aerial
observations by M.J. Hole, British Antarctic Survey) of Sims
Island suggest that it may be related to the Snow Nunataks
volcanic field and the Mount Benkert Formation.
Principal lithological characteristics:
The sequence exposed
at Mount McCann consists
of
200 m of pillow basalt capped
by
1-5
m
of
massive to well-bedded orange-brown
volcaniclastic rocks described
as
"hyaloclastite" (probably
comparable
to
the gravelly sandstones described for outcrops
on Beethoven Peninsula, Rothschild Island and Seal Nunataks).
By contrast, about 350 m
of
vertical rock face at Mount
Benkert is formed entirely of crudely bedded "hyaloclastite",
minor pillow lava and pillow breccia and thin massive lavas.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
374 J.L. SMELLIE
Smaller channels and cross-bedded layers are
also
recorded
and the outcrop is notable for the presence of spectacular,
multiple large-scale channel-like unconformities up to
50
m
deep. The sequence at Mount Thornton also resembles that at
Mount Benkert and is principally formed of massive and thin-
bedded volcaniclastic rocks, which overlie poorly exposed
pillowed and blocky lava; evidence for slope instability
(convolute layering, folding) is common. Basalt dykes, sills
and irregular small plugs are present at
all
three localities.
There is avery strong resemblance in lithofacies and lithofacies
relationships to the Mussorgsky Peaks and Bruce Nunatak
formations, although the Mount Benkert Formation has been
studied at reconnaissance level only. The outcrops are
tentatively interpreted
as
representing the pillow volcano and
subaqueous tuff cone stages of several small subglacial
volcanoes, erupted beneath relatively thick ice sheet(s).
Thickness:
More than 350 m are exposed at Mount Benkert
and Sims Island.
Boundaries:
Contact relationships with other rock formations
are unclear and the pre-volcanic basement is unexposed. The
formation is overlain at Mount McCann by subaerial lithofacies
of theMount McCann Formation, but the nature of thecontact
is undescribed. Elsewhere, the upper boundary corresponds
to the present-day erosion surface
Age:
Unknown, but alkaline composition, lithofacies
similarities and freshness suggest that it is part of the Miocene
to Recent volcanism of the Antarctic Peninsula region.
Type section:
Mount Benkert.
Key references:
O'Neill
&
Thomson (1985), Thomson
&
O'Neill (1990).
Mount McCunn Formation
Name:
After Mount McCann (73"34'S, 77"37'W), one of the
Snow Nunataks, rising to
c.
700 m and situated south of
Carroll Inlet, English Coast.
Distribution:
Snow Nunataks (Espenchied Nunatak; upper
sequence at Mount McCann) and Rydberg Peninsula (isolated
lava outcrop and possibly Mount Combs; Fig.
6).
Principal lithological characteristics:
Scoriaceous and cindery
basaltic rubble and massive highly vesicular lavas (Mount
McCann); crudely stratified, black and reddish-brown lapilli
tuff and tuff breccia intruded by thin
(1-3
cm) basaltic dykes
(EspenchiedNunatak). The tiny outcrop on Rydberg Peninsula
is formed of black highly vesicular lava. Mount Combs was
described
as
a
small cone that rises several hundred metres
above the surrounding ice surface. Although it is likely to be
volcanic, it is entirely snow and ice covered. All the outcrops
of the Mount McCann Formation are dominated by subaerial
lavas and possible Strombolian tephra. Those at Snow
Nunataks possibly represent the subaerial caps to subglacially
erupted volcanoes (cf. Mount Grieg and Christensen Nunatak
formations).
Thickness:
More than
60
m
at
Mount McCann.
Boundaries:
No contacts with other rock formations are
exposed. The formation overlies subaqueous lithofacies of the
Mount Benkert Formation at Mount McCann, but the contact
is unexposed. Elsewhere, the upper surface is the present-day
erosion surface.
Age:
Unknown, but alkaline composition, lithofacics
similarities and freshness suggest that it is part of the Mioccnc
to Recent volcanism of the Antarctic Peninsula region.
Type section:
Upper sequence at Mount Thornton (described
above).
Key references:
O'Neill
&
Thomson
(1983,
Smellie
et
al.
(1988),
Rowley
&
Thomson
(1990),
Thomson
&
O'Ncill
(1
990).
Henry Nunutuks Formation
Nume:
After Henry Nunataks (75"08'S, 72"36'W),
on
the
western side of the Merrick Mountains, eastern Ellsworth
Land.
Distribution:
Merrick Mountains only. Central nunatak
and
possibly western end
of
the easternmost nunatakin thc Hctwy
Nunataks; also an unnamed isolated nunatak
5
km west
of
Eaton Nunatak (Fig.
6).
Principal lithological characteristics:
Examined only
;it
reconnaissance level. The Henry Nunataks outcrop consisils
of a sequence
of
fine-grained, grey, vesicular basaltic
laws
with rubbly surfaces cut by
a
basalt dyke. At the small
nunatak
west of Eaton Nunatak is
a
frost-shattered basanite
la~a
breccia with apalagonite-altered glassy matrix (interpreted
;IS
hyaloclastite breccia) overlain by thinner vesicular to
scoriaceous lavas.
Thickness:
About
100
m at Henry Nunataks;
<I2
m
at
tl-lc
small nunatak west of Eaton Nunatak.
Boundaries:
The Henry Nunataks sequence unconformably
overlies Mesozoic volcanic rocks (mostly porphyritic dacitic
lavas). The base of the other outcrop west of Eaton
Nunatak
is obscured by ice but it is probably unconformable on Jurassic
sedimentary rocks widely exposed in the surrounding arca.
The upper surface is the present-day crosion surface.
Age:
A K-Ar age of 6 Ma was attributed to the outcrop nc;u
Eaton Nunatak, but neither analytical details nor the errors lor
the age were published.
Type section:
Central nunatak at Henry Nunataks.
Key References:
Halpern (1971), Smellie
et al.
(198X),
Thomson
&
Kellogg (1990), Rowley
et
al.
(1990).

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY
OF
MONOGENETIC VOLCANIC FIELDS 375
Discussion
rocks. The JRIVG volcanism was longer-lived than for any
Characteristics
of the
volcanic fields
The volcanic fields vary in size from about
I
to
4500
km2
(Table 11) and have
a
range of ages from early Miocene
(1
5
Ma) to Recent. The James
Ross
Island Volcanic Group is
probably the largest area of late Miocene-Recent volcanic
rocks preserved
in
the Antarctic Peninsula region, cropping
out over
c.
4500
km2.
It comprises a large polygenetic shield
volcano on James Ross Island (Mount Haddington), which is
35-60kmin basal diameter.
It
was largely effusive, constructed
from multiple superimposed hyaloclastite deltas, but includes
the products of several tuff cone centres (the largest, at
Terrapin Hill, may be
8
km in diameter). Other mainly
effusive shield volcanoes in the JRIVG are known or postulated
on Ulu Peninsula (major centre at Dobson Dome?), Tabarin
Peninsula (20-40 km in basal diameter), Vega Island and
islands
in
Antarctic Sound (Sykes 1988b, Smellie 1990a).
The elongate morphology of the Vega Island outcrops suggest
that the volcanic sequences there were erupted from a fissure
or series of fissures, unusual in the JRIVG. Most of the islands
in
Prince Gustav Channel are small monogenetic centres
constructed of alarge proportion ofpyroclastic (hydrovolcanic)
other volcanic outcrop in the region, and may exceed
10
million
years. The distribution of ages suggests that there may be a
simple age progression in the JRIVG, with youngercentres
(<
2
Ma) situated mainly to the north and north-east of the older
Mount Haddington centre (Baker
et
al.
1977, Sykes 1988b).
By contrast, other Miocene-Recent volcanic fields
in
the
Antarctic Peninsula are monogenetic and available data indicate
three principal eruptive periods
(c.
15,
7.7-5.4 and 2.7-
<0.2
Ma).
The lifetime
of
each monogenetic field was
typically <I-2 million years and some formations are single
short-lived volcanic centres (e.g. Argo Point and Hornpipe
Heights). The Hornpipe Heights Formation appears to
represent a resumption of volcanic activity
in
the Mount
Pinafore volcanic field, after
a
break of
3-3.5
million years.
With the possible exception
of
Seal Nunataks, there appear to
be no spatial or temporal controls on the distribution of the
centres within any monogenetic field, but available information
is
sparse. The isotopic ages also suggest that monogenetic
fields formed by lavas with basanite-phonotephrite
compositions are generally older than those with alkali basalt-
tholeiite lavas (Table 11).
There is a large disparity
in
the volume of erupted products
Table
11.
Summary statistics for alkaline volcanic fields in the Antarctic Peninsula and eastern Ellsworth Land, grouped according to petrological "series".
Volcanic field Approx. area Estimated preserved Age Formations Comments Reference
(km') thickness of deposits (m)
(Ma)
Alkali Basalt-Tholeiite "Series"
JIRVG 4500
Seal Nunataks 1400
Argo Point
[I1
Beethoven Peninsula >2000
(>7000?**)
Snow Nunataks 2800
Basanite-Phonotephnte "Series"
Mount Pinafore 250
Rothschild Island [51
Menick Mountains [I51
Venus Glacier*** 500
up to 1470 m;
individual outcrops
usually few hundred
metres
c.
300-650*
>I75
600-800*
500-600*
(likely
>>
600
m
at Sims Island)
20-90
100
12-100
0.1-2.3
<I0
(?4)-<1
0.8-1.6
2.5, 0.68
undated
7.7-6.0;
2.7-2.5
5.4
Hobbs Glacier,
Cockbum Island;
no volcanic
formations defined yet
Bruce Nunatak,
Christensen Nunatak
Argo Point
Mussorgsky Peaks,
Mount Grieg
Mount Benkert,
Mount McCann
Mount Pinafore,
Hornpipe Heights
Overton Peak
6
(1
locality) Henry Nunataks
IS
(1
dyke) Venus Glacier
includes several Nelson 1975,
large shield this paper
volcanoes
multiple overlapping Smellie
&
centres Hole 1997
single volcanic Saunders
centre
1982
multiple overlapping Smellie
&
centres Hole 1997
several scattered this paper
centres
few widely separated
centres 1993
two small very Care 1980
degraded outcrops
two small very this paper
degraded outcrops
dykes and sills only
Smellie
er
(I/.
this paper
*
assumes continuity of outcrop
to
sub-ice bedrock
**
note: the distribution of magnetic anomalies suggests that the Beethoven Peninsula volcanic field may be significantly larger than the distribution of
surface exposures suggests, possibly extending between Monteverdi Peninsula and Latady
Island
(Renner
et
uf.
1982)
***
compositionally heterogeneous; affinities uncertain
[
]
tentative estimates referring to small volcanic fields consisting of 1-2 centres only

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
376
J.L.
SMELLIE
across the region, which varies with the compositions of the
erupted magmas. Two broad petrogenetic “series” were
recognised by Smellie
(1
987) and form essentially mutually
exclusive outcrops. Basanite-phonotephrite outcrops contain
thin sequences (<lo0 m thick) erupted from small widely
separated centres scattered across comparatively small volcanic
fields
(<300
km2, Table 11). By contrast, eruptions of alkali
basalts-tholeiites were much more voluminous and occur
within much larger volcanic fields (>1500 km2). There is no
chronological progression between alkali and tholeiite basalts
in
any field, consistent with petrogenetic interpretations.
Fractional crystallization was generally minor and the broad
range of basalt types present is believed to be due to partial
melting variations (Hole 1988,1990, Sykes 1988a). Relative
degrees of partial melting probably decreased from the
tholeiites to alkali basalts. Eruptions were variously from
large shield volcanoes (JRIVG) or from multiple, closely
spaced and probably overlapping much smaller centres. The
largest alkali basalt-tholeiite volcanic fieldmay be the JRIVG
(c.
4500
km2), but the monogenetic field at Snow Nunataks is
about 2800
km2
and that
in
south-western Alexander Island
possibly extends over >7000 km2 based on aeromagnetic
surveys (Renner
et
al.
1982). Because all of the outcrops are
highly degraded,
it
is impossible to estimate accurately the
volumes of magma erupted. However, from a consideration
of the relative thicknesses of preserved deposits and the areal
extent of the volcanic fields, erupted volumes of alkali basalt-
tholeiite were an order
of
magnitude greater, at least, than
those of basanite-phonotephrite. This is true even if data for
the JRIVG polygenetic volcanic field are excluded.
Eruptive palaeoenvironments
Eruptive/depositional environments varied between marine
and englacial in the JRIVG (Nelson 1975, Skilling 1994).
During the glacial periods, features preserved
in
the volcanic
sequences suggest that coeval glaciers were at least 200 m
thick (cf. Smellie in press a, in press b). Conversely, most of
the volcanism
in
the monogenetic volcanic fields was englacial.
Eruptions were beneath valley-confined Alpine glaciers in
northern Alexandcr Island (Mount Pinafore Formation).
Interpretation of the lithofacies
in
the Mount Pinafore
Formation suggests that glaciers were mainly mainly composed
of snow and firn; glacier thicknesses were probably comparable
at each outcrop and a few tens of metres thick. The Mount
Pinafore Formation outcrops represent “outflow facies“ of
subglacial eruptions of “sheetflow type” beneath valley-
confined glaciers (Smellie
etal.
1993). The variable elevations
of the basal pre-volcanic basement at each locality
(1
100 m to
c.
400
m) suggest a significant bedrock palaeotopography and
sloping glacier surfaces. The present-day occurrence of most
of the outcrops capping ridges indicates that topographic
inversion has occurred following substantial post-Miocene
erosion. By contrast, products of eruptions beneath much
thicker glaciers (corresponding to ice sheets several hundred
metres thick) are represented by outcrops at Beethoven
Peninsula, Seal Nunataks (Smellie
&
Hole 1997) and probably
also the Overton Peak, Mount Bcnkert and Mount McCann
formations.
Eruptions
in
the Mount Pinafore and Rothschild Island
volcanic fields overlapped
in
time and occurred
in
geographically adjacent areas (Fig. 3). The sequences also
display features consistent with contrasting ice conditions
during the coincident eruptive period. Alpine (valley) glacierii
affected vents at higher elevations (Mount PinaforeFormation;
Smellie
et al.
1993) whereas an ice sheet was present
farther
west, at a lower elevation (affecting the Overton Peak
Formation). These conditions are not unlike those prevailing
at present but with generally higher ice surfaces in the Miocene.
It seems likely that the Miocene glacial topography was
broadly similar to that
of
today andreflecteddiffering bedrock
elevations between Rothschild Island and Mount Pinaforc.
Solely subaerial (“ice-free“) volcanism is recorded
by
the
nearby Hornpipe Heights Formation, yet the outcrop of
the
latter is at a similar topographical elevation to lower parts
of
the (englacial) Mount Pinafore Formation, indicating
that
between latestMiocene(c.
5.4Ma)andPliocenetimes(2.5Ma),
thegeneral elevation of glacicr surfaces
in
northern Alexander
Island diminished. As the base of the Hornpipe Heights
Formation is now covered by Sullivan Glacier, local glacicr
elevations may also have increased slightly since Plioccne
times, although
it
is not possible toestimate the magnitudes
of
those fluctuations.
A record of surface elevations of former ice sheets is
also
contained
in
the extensive JRIVG, Beethoven Peninsula and
Seal Nunataks outcrops, and probably also at Snow Nunataks
which currently lack diagnostic information. Although thc
JRIVG outcrop contains the
most
comprehensive record
oi‘
temporally fluctuating ice surfaces, the distinction between
eruptiveunits formed under marine andenglacial environments
is unclear at present and limits its usefulness for understanding
the palaeoenvironmental history of the Antarctic Peninsula.
Conversely, for the monogenetic volcanic fields at Beethoven
Peninsula and Seal Nunataks (and probably Snow Nunataks),
the Pliocene-Recent ice sheets were substantially thicker
and
had surface elevations typically a few hundred metres higher
than at present (Smellie
&
Hole 1997). These outcrops are
now extensively glacially degraded and several show evidcncc
for overtopping by former ice. The evidence thus clearly
indicates that ice sheet elevations on the low-lying flanks
of
the Antarctic Peninsula have fluctuated substantially (within
a few hundred metres, at least) over the past 2.5 million years,
before diminishing to those of the present-day.
Acknowledgements
The author is grateful to Dr M.R.A. Thomson, A. Giret
and
P.H.H. Nelson for comments on this paper.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
LITHOSTRATIGRAPHY
OF
MONOGENETIC
VOLCANIC
FIELDS
377
References
ADIE, R.J. 1953.
The rocks
of
Gruhum Lund.
Ph.D. thesis, University
of Cambridge, 259 pp. [Unpublished].
AITKENHEAD, N. 1975. The geology
of
the Duse Bay-Larsen Inlet area,
north-east Graham Land.
British Anturctic Survey Scientific Reports,
No. 5
I,
62 pp.
ANDERSON,
J.G. 1906. On the geology of Graham Land.
Bulletin
of
the Geological Institution
rfl
the University
of
Uppsulu,
7,
19-7
1.
BAKER, P.E., BUCKLEY,
F.
&
REX, D.C. 1977. Cenozoic volcanism in
the Antarctic.
Philosophical Trunsuctions
(?f
the Royul Society
of
London,
B279,
131-142.
BARKER, P.F. 1982. The Cenozoic subduction history of the Pacific
margin
of
the Antarctic Peninsula: ridge crest-trench interactions.
Journul
of
the Geologicul Society, London,
139,
787-801.
BELL, C.M. 1973. The geology
of
Beethoven Peninsula, south-western
Alexander Island.
British Anturctic Survey Bulletin,
No.
32.75-83.
BIBBY, J.S. 1966. The stratigraphy of part
of
north-east Graham Land
and the James Ross Island group.
British Antarctic Survey Scientific
Reports,
No. 53, 37 pp.
BORGIA, A. 1994. The dynamic basis of volcanic spreading.
Journul
of
Geophysicul Reseurch,
99,
17791-17804.
BURN, R.W.
&
THOMSON, M.R.A. 198
I.
Late Cenozoic tillites associated
with intraglacial volcanic rocks. Lesser Antarctica.
In
HAMBREY,
M.J.
&
HARLAND, W.B.,
eds. Pre-Pleistocene tillites:
u
record
of
Earth’s
gluciul
history.
Cambridge: Cambridge University Press,
199-203.
CARE, B.W. 1980. The geology of Rothschild Island, north-west
Alexander Island.
British Antarctic Survey Bulletin,
No. 50, 87-1 12.
CRAME, J.A., PIRRIE, D., RIDING, J.B.
&
THOMSON, M.R.A. 1991.
Campanian-Maastrichtian (Cretaceous) stratigraphy of the James
Ross Island area, Antarctica.
Journal
of
the Geologicul Society,
London,
148,
1125-1 140.
DEL
VALLE, R.A.
&
MEDINA, F.A. 1985. Geologia de Cabo Marsh,
Isla
Robertson, Antartida.
Contribucicin
Institutu
Antdrtico Argentino,
DEL
VALLE, R.A.
&
RINALDI, C.A. 1992. Regional scheme
of
the main
structural features
of
the northeastern extreme
of
the Antarctic
Peninsula and the James Ross Island area.
In
RINALDI, C.A.,
ed.
Geologiu de
la
islu
Jumes Ross.
Buenos Aires: Instituto Antirtico
Argentino, 349-358.
DEL
VALLE, R.A.,FOURCADE, N.H. &MEDINA, F.A. 1983. Interpretacih
preliminar de
las
edades K-Ar y de
10s
analisis quimicos de
las
rocas
volcinicas y de
10s
diques de
10s
nunataks Foca, Antartida.
Contribuciones del Institutu Antdrtico Argentino,
No.
287,
13
pp.
1997. Jurassic trees at Jason Peninsula, Antarctica.
Anturctic
Science,
9,
443-444.
DEL
VALLE, R.A., MARSHALL, P.A., LIRIO, J.M.
&
CAMERLINGO,
E.
1987.
Sobre
la
presencia del conglomerado con pecten en
el
fiordo Beltn,
isla James Ross.
Resimenes, Primeru Reunion de Communicuciones
sobre Investiguciones Antdrticus, Buenos Aires,
16
iil
20
de
Noviembre de
1987.
Buenos Aires: Direccion Nacional de Antirtico,
I
p. (no pagination)
DINGLE, R.V.
&
LAVELLE, M. 1998. Antarctic Peninsular cryosphere:
Early Oligocene
(c.
30 Ma) initiation and arevised glacial chronology.
Journul
of’
the Geological
Society,
London,
155,
433-437.
FLEET, M. 1968. The geology of the Oscar
11
Coast, Graham Land.
British Anturctic Survey Scientific Reports,
No. 59, 46 pp.
GONZALEZ-FERRAN,
0.
1983. The Seal Nunataks: an active volcanic
group on the Larsen Ice Shelf, West Antarctica.
In
OLIVER, R.L.,
JAMES, P.R.
&
JACO, J.B.,
eds. Antarctic eurrh science.
Canberra:
Australian Academy of Science, 334-337.
GRUNOW, A.M. 1993. New paleomagnetic data from the Antarctic
Peninsula and their tectonic implications.
Journcil
of
Geophysical
Reseurch,
98,
13 815-13 833.
309,
1-29.
DEL
VALLE, R.A., LIRIO, J.M., LUSKY, J.C., MORELLI, J.R.
&
NUNEZ, H.J.
HALPERN,
M.
I97
I.
Evidence for Gondwanaland from
a
review of West
Antarctic radiometric ages.
In
QUAM, L.O.,
ed. Reseurch
in
the
Antarctic.
Washington, DC: American Association for the
Advancement of Science, 717-730.
HOLDSWORTH, B.K.
&
NELL, P.A.R. 1992. Mesozoic radiolarian faunas
from the Antarctic Peninsula: age, tectonic and palaeoceanographic
significance.
Journal
of
the Geologicul Society, London,
149,
1003-1020.
HOLE, M.J. 1988. Post-subduction alkaline volcanism along the
Antarctic Peninsula.
Journul
of
the Geological Society, London,
145,
985-988.
HOLE, M.J. 1990a. Geochemical evolution of Pliocene-Recent post-
subduction alkalic basalts from Seal Nunataks, Antarctic Peninsula.
Journal
of
Volcunology cind Genthermul Reseurch,
40,
149-167.
HOLE, M.J. 1990b. Beethoven Peninsula.
Anrurctic Reseurch Series,
HOLE, M.J. I990c. Hornpipe Heights.
Anturctic Reseurch Series,
48,
HOLE, M.J.
&
SAUNDERS, A.D. 1996. The generation of small melt-
fractions in truncatedmelt columns: constraints frommagmaserupted
above slab windows and implications for MORB genesis.
Minerulogicul Muguzine,
60,
173-1 89.
HOLE, M.J. &THOMSON, J.W. 1990. Mount Pinafore-Debussy Heights.
Anturctic Reseurch Series,
48,
268-270.
HOLE, M.J., KEMPTON, P.D.
&
MILLAR, 1.L. 1993. Trace element and
isotope characteristics
of
small degree melts
of
the asthenosphere:
evidence from the alkalic basalts of the Antarctic Peninsula.
Chemical
Geology,
109,
51-68.
HOLE, M.J., ROGERS, G., SAUNDERS,
A.D.
&
STOREY,
M.
1991a. The
relationship between alkalic volcanism and
slab
window formation.
Geology,
19,
657-660.
HOLE, M.J., SMELLIE, J.L.
&
MARINER, G.F. I991 b. Geochemistry and
tectonic setting of alkali-basalts from Alexander Island.
In
THOMSON,
M.R.A., CRAME, J.A.
&
THOMSON, J.W.,
eds. Geologicul evolution
of
Anturcticu.
Cambridge: Cambridge University Press, 521-526.
HOLE, M.J.H., SAUNDERS, A.D., ROGERS, G.
&
SYKES, M.A. 1995. The
relationship between alkaline magmatism, lithospheric extension
and slab window formation along continental destructive plate
margins.
In
SMELLIE, J.L., ed.
Volcanism ussociufed
with
extension
at
consuming plcite murgins.
Geological Society, London, Special
Publication,
No.
8
I,
265-285.
HORNE, R.R.
&
THOMSON, M.R.A. 1967. Post-Aptian camptonite dykes
in south-east Alexander Island.
Briti.sh Anturctic Survey Bulletin,
No. 14, 15-24.
JONKERS,
H.A. 1998a. The Cockburn Island Formation;
late
Pliocene
interglacial sedimentation in the James Ross Basin, northern Antarctic
Peninsula.
Newsletter on Strutigruphy,
36,
63-76.
JONKERS, H.A. I998b. Stratigraphy of Antarctic late Cenozoic pectinid-
bearing deposits.
Anfurctic Science.
10,
161-170.
JONKERS, H.A.
&
KELLEY, S.P. 1998. A reassessment of the
age
of the
Cockburn Island Formation, northern Antarctic Peninsula, and its
palaeoclimatic implications.
Journul
of
the Geological Society,
London,
155,
737-740.
LARTER, R.D. 1991. Debate: Preliminary results of seismic reflection
investigations and associated geophysical studies in the area of the
Antarctic Peninsula.
Antiirctic Science,
3,
217-220.
LAWVER,L.A.,KELLER,R.A.,FISK,M.R.
&STRELIN,J.A.
1995.
Bransfield
Strait, Antarctic Peninsula. Active extension behind
a
dead arc.
fn
TAYLOR, B.,
ed. Buckurc busins: tectonics und mcigmutism.
New
York: Plenum Press, 3 15-342.
LE BAS, M.J., LE MAITRE, R.W., STRECKEISEN, A.
&
ZANETTIN, B. 1986.
A chemical classification of volcanic rocks based on the total alkali-
silica diagram.
Journul
of
Petrology,
27,
745-750.
LEMASURIER, W.E.
&
THOMSON, J.W.,
eds.
1990. Volcanoes of the
Antarctic plate and southern oceans.
Antarctic Reseurch Series,
48,
487 pp.
48,
273-276.
21 1-272.

http://journals.cambridge.org Downloaded: 16 Dec 2013 IP address: 221.130.17.50
378
J.L. SMELLIE
MALAGNINO, E.C., OLIVERO, E.B., RINALDI, C.A.
&
SPIKERMANN, J.P.
1978.
Aspectos geologicos del horde occidental de
la
Isla James
Ross, Antirtida.
In Actcis del VII Congres GeolAgico Argentino,
NeuquPn,
1,
Buenos Aires: Asociacion Geologica Argentina,
489-503.
MASSABIE, A.C.
&
MORELLI, J.R.
1977.
Buchitasde IaIslaVicecomodoro
Marambio, sector Anthrtico Argentino.
Revi.rtci de
la
Asociacicin
Geoldgicu Argentinu,
32,
44-5
I.
MCCARRON. J.J.
&
LARTER, R.D.
1998.
Late Cretaceous to early
Tertiary subduction history of the Antarctic Peninsula.
Journal
of’
the Geologicul Society, London,
155,
255-268.
MCCARRON, J.J.
&
MILLAR, I.L.
1997.
The age and stratigraphy of fore-
arc magmatism on Alexander Island, Antarctica.
Geological
Muguzine,
134, 507-522.
MCCARRON, J.J.
&
SMELLIE. J.L.
1998.
Tectonic implications of fore-
arc magmatism and generation
of
high-magnesian andesites:
Alexander Island, Antarctica.
Journul
of
the Geologicul Society,
MONCRIEFF, A.C.M.
&
KELLY, S.R.A.
1993.
Lithostratigraphy of the
uppermost Fossil Bluff Group (Early Cretaceous) of Alexander
Island, Antarciica: history
of
an Albian regression.
Crettic.eou.s
Keseurch.
14,
1-15.
NELSON,
P.H.H.
1975.
The James Ross Island Volcanic Groupof norih-
east Graham Land.
British Antnrcfic Survey Scienfific Reports,
No.
54. 62
pp.
O’NEILL, J.M.
&
THOMSON, J.W.
1985.
Tertiary mafic volcanic and
volcaniclastic rocks of the English Coast, Antarctica.
Anturctic
Journcil
of
the United Stutes,
20(5),
36-38.
PIRRIE,
D.
&
SYKES, M.A.
1987.
Regional significance of proglacial
delta-front reworked tuffs, James Ross Island.
British
Anturctic
Survey Bulletin,
No.
77, 1-12.
PIRRIE,
D.,
CRAME, J.A., RIDING. J.B.
&
TAYLOR, P.D.
1997.
Miocene
glaciomarine sedimentation in the northern Antarctic Peninsula
region: the stratigraphy and sedimentology
of
the Hobbs Glacier
Formation, James Ross Island.
Geologicul Magcizine,
136,745-762.
RENNER, R.G.B., DIKSTRA, B.J.
&
MARTIN. J.L.
1982.
Aeromagnetic
surveys over the Antarctic Peninsula.
In
CRADDOCK, C.,
ed. Antcirctic
geoscience.
Madison, W1: University of Wisconsin Press,
363-370.
REX,
D.C.
1970.
Age of
a
camptonite dyke from south-east Alexander
Island.
British Antcirctic Survey
Bulletin,
No.
23.
103.
REX, D.C.
1976.
Geochronology in relation to the stratigraphy of the
Antarctic Peninsula.
Britislt Anturctic Survey Bulletin,
No.
43,
Ricci. C.A., HERVE,
F.,
KRYNAUW, J.R.
&
LEMASURIER. W.E.
1993.
Naming of igneous and metamorphic rock units in Antarctica:
a
recommendation by thc SCAR Working Group on Geology.
Antarctic
Science,
5,
103-
104.
RILEY, T.R., CRAME, J.A., THOMSON, M.R.A.
&
CANTRILL, D.J.
1997.
Late Jurassic
(Kimmeridgian-Tithonian)
macrofossil assemblage
from Jason Peninsula, Graham Land: evidence for a significant
northward extension of the Latady Formation.
Anturctic
Science,
9,
434-442.
ROWLEY, P.D.
&
SMELLIE, J.L.
1990.
Southeastern Alexander Island.
Anturctic Reseurch Series,
48, 277-279.
ROWLEY, P.D.
&
THOMSON, J.W.
1990.
Rydberg Peninsula.
Antarctic
Reseurch Series,
48, 280-28
I.
ROWLEY, P.D.,VENNUM, W.R.
&
SMELLIE, J.L.
1990.
MerrickMountains.
Anturctic Rrseurch Series,
48, 296-297.
SAUNDERS, A.D.
1982.
Petrology and geochemistry of alkali-basalts
from Jason Peninsula, Oscar I1 Coast, Graham Land.
British Antcirctic
Survey Bulletin,
No.
55,
1-9.
SKILLING, I.P.
1994.
Evolution of an cnglacial volcano: Brown Bluff,
Antarctica.
Bullefin
of
Volcunology,
56, 573-59
1.
L(lndon,
155, 269-280.
49-58.
SMELLIE, J.L.
1987.
Geochemistry and tectonic setting of alkalinc
Journul
01
SMELLIE, J.L.
1990a.
Seal Nunataks.
Anturctic Reseurch Serre.s,
48.
SMELLIE, J.L.
1990b.
Graham Land and South Shetland Islands
summary.
Anturctic
Research Series.
48,
303-3
12.
SMELLIE, J.L.
199
I.
Middle-Late Jurassic volcanism on Jason Peninhula,
Antarctic Peninsula, and its relationship to the break-up of Gondwana
In
ULBRICH,
H.
&
ROCHA CAMPOS, A.C.,
eds. Gorrdwunci
Srveit
proceedings.
Slo
Paulo: Universidade de
S%o
Paulo.
685-699.
SMELLIE, J.L. In press a. Lithofacies architecture and construction
ol
volcanoeserupted inenglacial lakes: tcefall Nunatak. Mount Murphy,
eastern Marie Byrd Land, Antarctica.
In
Whim, J.D.R.
&
Rims,
N
.
e ds.
La
cus
f
r
in
e
v
o
1
cu
n
i
clus
ti
c,
.s
r
dim
en
tu
tio
n.
Intern at
i
o
11
a1
Association of Sedimentologists Special Publication.
SMELLIE, J.L. In press b. Subglacial eruptions.
In
SicuRossoN,
H.,
ed
Encyclopediei
of
volcunoes.
San Diego: Academic
Press.
SMELLIE. J.L.
&
HOLE, M.J.
1997.
Products and processes in Plioccnc
-
Recent, subaqueous to emergent volcanism in the Antarctic Peninsula:
examples of englacial Surtseyan volcano construction.
Bulletin
(I/
Vulcunology,
58, 628-646.
SMELLIE, J.L.
c%
SKILLING. I.P.
1994.
Products of subglacial eruptions
under different ice thicknesses: two examples from Antarclica.
Sedimentury Geology,
91,
1
15-1
29.
SMELLIE, J.L.. HOLE, M.J.
&
NELL, P.A.R.
1993.
Late Miocene
valley^
confined subglacial volcanism in northern Alexander Island, Antarctic
Peninsula.
Bulletin
of
Volcanology,
55,
273-288.
1988
Age, distribution and eruptive conditions of late Cenozoic alkalint
volcanism in the Antarctic Peninsula and eastern Ellsworth Land
British Antarctic Survey Bulletin,
No.
80,
21-49.
STOREY. B.C., VAUGHAN. A.P.M.
&
MILLAR, I.L.
1996.
Geodynarnic
evolution
of
the Antarctic Peninsula during Mesozoic times and
IIS
bearing on Weddell Sea history.
In
STOREY, R.C.,
KING,
E.C
&.
LIVERMORE, R.A.,
ed.7. Weddell Seti tectonics und
Gondwunu
hrrtrk
up.
Geological Society, London, Special Publication, No
I
OX.
STRELIN,
J.
&
MALAGNINO, E.C.
1992.
Geomorfologia de la isla Jamcc.
Ross.
In
RINALDI, C.A.,
ed. Geologiu de
I(i
isla
James Ross.
Buenw,
Aires: Instituto Antirtico Argentino,
7-36.
SYKES, M.A.
I98Xa.
The
petrology
rind
tectonic
.sign<fi’cancc
of
rhl,
Jurnes Ross Island Volccinic. Group, Antarctica.
Ph.D. thcslb.
University of Nottingham. [Unpublished].
SYKES, M.A.
1988b.
New K-Ar age determinations on thc James
Ro?\
Island Volcanic
Group,
north-east Graham Land. Antarctica.
Urrr,
r/r
Antarc,tic. Survey Bulletin,
No.
80,
5
1-56.
TIIOMSCIN, J.W.
1990.
Argo Point.
Anturclic Research Sertr.7,
4X.
352-353.
THOMSON, J.W.
&
KELLOGG,
K.S.
1990.
Henry Nunataks.
Autul-,
/J;
Reseurch Series,
48, 294-295.
THOMSON, J.W.
&
O’NEILL, J.M.
1990.
Snow Nunataks.
Anitrr~i,,
Research Series.
48, 283-285.
VAN
WYK
DE
VRIES, B.
&
BORGIA, A.
1996.
The role of basement
iii
volcano deformation.
In
MCGUIRR, W.J., JONES, A.P.
&
NEUHFKG,
J
,
eds. Volccino instability
on
the Earth und other
plnnets.
Geological
Society, London, Special Publication, No.
110,
95-1
1
I.
WHITTAKER, A., COPE, J.C.W., COWIE, J.W., GIBBONS, W., HAII.WOO~~.
E.A., HOUSE, M.R., JENKINS,
D.G..
RAWSON,
P.F., RUSHTON, A
W
I..
SMITH,
D.G.,
THOMAS, A.T.
&
WIMBLEDON, W.A.
1991.
A guide
to
stratigraphical procedure.
Journrrl
of
the Geologicul Society, /.ondon,
148.
8
13-824.
volcanic rocks in the Antarctic Peninsula: a review.
Volcunology and Geothermal Research,
32,
269-285.
349-35
I.
SMELLIE, J.L., PANKHURST, R.J., HOLE, M.J.
&
THOMSON, J.W.
87-104.