GeolCarp_Vol58_No6_541

GEOLOGICA CARPATHICA, DECEMBER 2007, 58, 6, 541—563

www.geologicacarpathica.sk

Miocene volcanism in the Visegrád Mountains (Hungary):

an integrated approach to regional volcanic stratigraphy

DÁVID KARÁTSON

, ISTVÁN OLÁH

, ZOLTÁN PÉCSKAY

, EMŐ MÁRTON

, SZABOLCS

HARANGI

, ALFRÉD DULAI

, TIBOR ZELENKA

and SZABOLCS KÓSIK

Eötvös University, Department of Physical Geography, Pázmány Péter sétány 1/c, 1117 Budapest, Hungary; dkarat@ludens.elte.hu

Eötvös University, Department of Petrology and Geochemistry, Pázmány Péter sétány 1/c, 1117 Budapest, Hungary

Nuclear Research Institute, Hungarian Academy of Sciences, Bem József tér 18/c, 4001 Debrecen, Hungary

Eötvös Loránd Geophysical Institute, Paleomagnetic Laboratory, Columbus út 17—23, 1145 Budapest, Hungary

Hungarian Natural History Museum, Department of Geology and Paleontology, Ludovika tér 2, 1083 Budapest, Hungary

Hungarian Geological Survey, Stefánia út 14, 1143 Budapest, Hungary

In memoriam late László Korpás

(Manuscript received April 24, 2006; accepted in revised form March 15, 2007)

Abstract: A combined volcanological, petrographical, paleontological, radiometric and paleomagnetic approach to
the Middle Miocene volcanism of the Visegrád Mountains, Hungary, constrains their eruptive activity. The volcanic
evolution is divided into three stages on the basis of paleomagnetic data. The 1st

stage andesitic to dacitic explosive

eruptions occurred in a submarine environment  16 Ma. Their products show normal polarity and CCW declinations.
Scattered extrusive dacitic activity closely followed  16 Ma into a reverse polarity stage. A more voluminous,
subaerial, andesitic activity (16—15 Ma) produced a large variety of volcaniclastic rocks, mostly block-and-ash- and
debris-flow breccias, and lava domes/flows accompanied by subvolcanic bodies. The main eruptive centre was the
Keserűs Hill explosive lava dome complex. Within the andesitic activity, that partly overlapped the dacitic extrusive
activity, two paleomagnetic stages can be defined. These indicate a CCW rotation within a reverse polarity stage, 15.5—
15.3 Ma ago according to K/Ar data. All of the three paleomagnetic stages can also be found in the neighbouring
Börzsöny Mts. At the same time, the latter is characterized by a subsequent final stage (with no rotation in the High
Börzsöny lava dome complex) that is missing in the Visegrád Mts. With respect to paleomagnetism and the sporadic,
uncertain K/Ar data  < 14.5 Ma, in the Visegrád Mts the main andesitic volcanism may have terminated 15—14.5 Ma.
In the Börzsöny Mts, the buildup of the High Börzsöny (14.5—13.5 Ma) may have been coeval with only sporadic late-
stage andesitic eruptions in the Visegrád Mts.

Key words: Badenian, paleomagnetism, volcanology, K/Ar geochronology, submarine to emergent activity, lava
dome/flow complexes.

1. Introduction and previous work

In the past years, the application of new combined methods
(correlation of volcanic units with paleomagnetic, K/Ar
geochronological, geochemical as well as structural geo-
logical and paleontological data) have contributed to a
substantial refinement of the eruptive history and de-
tailed stratigraphy of the Miocene volcanic fields of the
North Hungarian Mountains, all belonging to the calc-al-
kaline Inner Carpathian Volcanic Chain. As a result, the
chronological evolution and stratigraphy of volcanism
from the Börzsöny Mts (Karátson et al. 2000) through the
Cserhát Mts (Póka et al. 2004) to the Bükk Foreland
(Márton & Pécskay 1998) could be much more precisely
constrained.

The Visegrád Mts are located in North Hungary

(Fig. 1: 250 km

) and belong to the initial calc-alkaline

volcanism of the Western Carpathians (e.g. Lexa &
Konečný 1974; Konečný & Lexa 1994; Szabó et al.
1992). Together with the northern-lying Börzsöny Mts,
they have long been regarded as a single, large Middle

Miocene volcanic field (e.g. Korpás (Ed.) 1998 and refer-
ences therein), composed of a number of lava domes and
small stratocones (Karátson 1995; Korpás (Ed.) 1998;
Harangi et al. 1999; Karátson et al. 2000; Karátson &
Németh 2001). However, whereas the volcanic stratigra-
phy of the Börzsöny Mts (with an initial explosive sub-
marine stage of 16—16.5 Ma and a late-stage subaerial
dome complex of 13.5—14.5 Ma: Karátson et al. 2000) is
well-defined, very few and uncertain data have been pub-
lished about the eruptive activity of the Visegrád Mts. A
great number of K/Ar datings were performed in the
1970’s by K. Balogh (1977—1979), but they remained
unpublished, and were not integrated with coeval paleo-
magnetic measurements (Balla & Márton-Szalay 1979).

In a historical view, the most important geological

contributions to the Visegrád Mts (e.g. Koch 1877;
Lengyel 1953; Zelenka 1960; Balla et al. 1977; Korpás
(Ed.) 1998; Harangi et al. 1999) agree about the exist-
ence of two fundamental rock associations: a garnet-
bearing biotite dacite lava and volcaniclastic unit, and a
predominantly amphibole andesite mostly volcaniclastic

542

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

unit. Stratigraphical constraints (e.g. Zelenka 1960) have
long indicated that the former should have preceded the
latter. On the other hand, the suggested duration of vol-
canic activity has been controversial. Pécskay et al.
(1995), on the basis of the available K/Ar ages, argued
for a 3 Myr long volcanism (16.5—13.5 Ma), whereas Kor-
pás (Ed.) 1998), grouping and averaging the age data, ad-
vocated to a much shorter activity (15.2—14.6 Ma).

The first, dacitic/rhyodacitic association can be found

mostly in the S periphery of the Visegrád Mts (Zelenka
1960; Balla et al. 1977; Korpás (Ed.) 1998), where it is
separated from Triassic limestone hills to the S by a Mio-
cene fault system (e.g. Fodor et al. 1999). The uniform
highest elevations of the volcanic and carbonate rocks
with flat denudation surfaces (Láng 1955), as well as
faults crossing both of them (e.g. Fodor et al. 1999;
Székely & Karátson 2004), argue for the intense erosion
and Miocene to Pleistocene disintegration of the whole
area. The garnet-bearing rock association consists of (1)
pumiceous pyroclastic and resedimented volcaniclastic
deposits (e.g. in the ravine of Holdvilág-árok, one of the

key localities) and (2) exhumed subvolcanic to deeply
eroded extrusive bodies (e.g. Csódi Hill, Lom Hill). In
some basal strata of the volcaniclastic deposits, the fossil
content indicates shallow-marine facies conditions prior
to and during the first volcanic eruptions (e.g. Koch
1877; Méhes 1941; Bohn-Havas & Korecz-Laky 1980),
similar to the earliest submarine activity of the neigh-
bouring Börzsöny Mts (Karátson et al. 2000; Karátson &
Németh 2001). This is in accordance with the early con-
clusions of Koch (1877), and the observations of Wein
(1939), who noted that the tuffs were emplaced firstly in
a subaqueous environment, later subaerially.

Shallow submarine volcanism in the Burda (Helemba)

Mts of South Slovakia, on the NW periphery of the Vi-
segrád Mts, was also pointed out by Konečný & Lexa
(1994). They reconstructed (1) viscous magma extrusion
in a submarine environment, accompanied by extensive
brecciation, alteration and hyaloclastite formation, (2)
explosions resulting in submarine pumice flows, and (3)
secondary (reworking) processes such as gravity sliding
and slumping of breccias as well as submarine debris

Fig. 1. Shaded relief map (based on the 1: 50,000 DEM of Hungary) and geographic setting of the Visegrád Mts.

Fig. 2. Simplified volcanological map of the Visegrád Mts, on the basis of the authors’ mapping and the maps of Koch (1877),
Schafarzik & Vendl (1929), Zelenka (1960), Balla et al. (1977), and Korpás & Csillag-Teplánszky (1999).

544

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

flows. Presence of garnet in the Burda rocks suggests an
early age in terms of the volcanic activity of the Viseg-
rád Mts.

In some peripheral areas, an andesitic phreatomagmat-

ic series is the initial volcanic product both to the N (Kor-
pás et al. 1967) and the S-SE part of the mountains
(Bendő et al. 2001; Karátson et al. 2006). In addition,
pumiceous, andesitic lithic-rich volcaniclastic deposits,
frequently interbedded with massive volcanogenic sand-
stone to claystone layers, are also widespread in some
basal successions (Karátson et al. 2006 and this paper).
The andesitic lithology as well as ca. 16 Ma K/Ar ages of
some andesite lavas (see later) indicate that the andesite
volcanism started simultaneously with the dacitic one,
although more locally.

Traditionally, the second volcanic unit in the Visegrád

Mts has been described as a widespread amphibole
andesitic association, occurring in the central and north-
ern part. Most rocks are volcaniclastics, but a number of
small- to medium-sized lava domes/flows and subvolca-
nic bodies are also exposed. The termination of the
andesitic volcanism is marked by the presence of overly-
ing, calcareous algae- and mollusc-bearing “Leitha”
limestone in the N (Fig. 2; Schafarzik & Vendl 1929),
and by two small calcareous algae-bearing occurrences
of pumiceous “tuffite” in the SE periphery (Koch 1877;
Zelenka 1960). The age of the limestone is Early Bade-
nian (Müller 1984; Dulai 1996).

As for volcanic source areas, it has been widely accept-

ed that most of the subvolcanic and/or extrusive rocks
correspond to vent areas (e.g. Zelenka 1960; Korpás (Ed.)
1998). In contrast, explanations for the dominant volca-
niclastic rocks have been missing or poorly constrained.
The lower unit, namely the dacitic pumiceous “stratovol-
canic” series has not been connected to any paleovolca-
nic centres. For the upper unit, that is the amphibole
andesitic series, a Somma-type structure was proposed by
Schafarzik & Vendl (1929) and Cholnoky (1937), agreed
by Balla et al. (1977) and recently by Korpás (Ed.) (1998).
Firstly Cholnoky described the “calderas” of the outer,
morphologically poorly defined, interrupted Dobogókő
Hill and the inner, semicircular Keserűs Hill rims, but no
caldera-forming mechanisms (e.g. related pyroclastic de-
posits or subsidence) have been documented so far.

Karátson et al. (2001, 2006) have identified the

Keserűs Hill volcano as the main, central lava dome com-
plex of the andesitic activity, destroyed northward by re-
peated dome and sector collapse events. It was shown
that deposits of high-energy mass flows (debris flows, de-
bris avalanches, e.g. the lower sequences of Szent Mihály
Hill and the upper part of Visegrád Castle Hill) can be in-
ferred N of the half-open “caldera”, which is a remnant of
a deeply eroded horseshoe-shaped depression. In addi-
tion, to the S, volcano-sedimentary facies changes (see
also in point 3.3) as well as identical lithology to the
central part of Keserűs Hill suggest the vicinity of
Dobogó-kő to be the distal facies of the Keserűs Hill vol-
canic edifice, being simply an upthrown segment of a
fault. (The fault itself was mentioned first by Láng

(1955), and indicated on map by Balla et al. (1977).)
Modifying and refining the maps of Balla et al. (1977)
and Korpás & Csillag-Teplánszky (1999), Karátson et al.
(2006) emphasized the role of NW-SE, N-S and less fre-
quent NE-SW striking faults dissecting the volcanic
structures.

From the late 1990’s, new research projects on the Vi-

segrád Mts have made it possible to produce a great
number of new K/Ar datings and paleomagnetic measure-
ments. Results of new geochemical analyses have also
been published (Harangi 1999; Harangi et al. 2001). A
paleogeographic study has focused on the Danube Bend
and its volcano-geomorphic background (Karátson et al.
2006). In this paper, we present a stratigraphical synthe-
sis based on new volcanological and petrological data,
as well as previous and new K/Ar geochronological,
paleomagnetic, geochemical and paleontological results.
In the light of the newly established stratigraphy of the
neighbouring Börzsöny Mts (Karátson et al. 2000), we
make a geochronological comparison, and we think that
the eruptive activity of both areas can be better revealed
due to the integrated approach.

2. Analytical techniques

The rock samples were subjected to regular polarizing

microscope investigation with a Nikon Labophot 2 mi-
croscope equipped with a Nikon Coolpix 4500. We used
standard 30 m-thick polished thin sections. Major and
trace element analysis of 45 samples from the volcanic
suite was carried out partly in the Geochemical Laborato-
ries of the Royal Holloway University of London (for
analytical details see Harangi et al. 2001) and partly in
the ACME Laboratories in Toronto (www.acmelab.com).
In the ACME Laboratory, both the major and trace ele-
ments were determined by ICP-MS techniques. The two
laboratories gave consistent results tested by analysing
the same samples at each site.

For K/Ar whole-rock analysis, a piece of rock approxi-

mately 2 kg in weight was collected, macroscopically
free of xenoliths and obvious alteration, and then in-
spected by thin-section. These samples were crushed and
sieved to 250—100 m, then washed and dried at 110 ºC
for 24 h. A portion of the dried fraction was ground re-
sulting in powder which was analysed for potassium. To
evaluate the reliability of K/Ar ages, various mineral sep-
arates (biotite, hornblende, feldspar, volcanic glass) have
also been dated. Incipient weathering of mineral sepa-
rates was checked for by electronprobe microanalysis
and X-ray diffraction. Amphibole, plagioclase and glassy
matrix phases were separated by common heavy liquid
(Sodium-Politungstite) and magnetic separation tech-
niques from the 0.063—0.125 and 0.0125—0.250 mm frac-
tions after crushing the samples. Final cleansing of the
phases was done by handpicking. Rocks containing
glass commonly yield rejuvenated ages because glass is
normally enriched in potassium and often hydrated and
submicroscopically devitrified, which results in the loss

545

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

of radiogenic argon, even at low temperature. As a conse-
quence, the argon loss may be related to the degree of
hydration. On the other hand, it is proven that anhydrous
glass in some cases can quantitatively preserve the ar-
gon. Therefore, in our work, the dating of pure glass sep-
arates aimed to determine the mininum age of the last
eruptive event on the study area.

The techniques of argon and potassium measurements

were similar to those described in Pécskay & Molnár
(2002). To determine potassium content approximately
0.1 g of grounded samples were digested in HF by adding
sulphuric and perchloric acids. Potassium concentrations
were measured by flame photometer using Li internal
standard. Depending on K content of the sample, weights of
0.1—1 g were used for argon extraction. The argon isotopic
composition of the purified argon was measured using an
isotopic dilution method and

Ar spike in a sector-type

mass spectrometer, operating in a static mode with a single
collector system. The result of calibration of the instruments
and the details of the applied methods have been described
elsewhere (Balogh 1985). Calculation of K/Ar ages was
done using the decay constants suggested by Steiger &
Jäger (1977). Analytical errors represent one standard
deviation. Interlaboratory standards Asia 1/65 LP-6, HD-B1
and GL-0 as well as atmospheric argon were used to control
the measurements (Odin 1982). In the stratigraphic evalua-
tion, we refer to the internationally accepted time scale
established by Vass & Balogh (1989).

The paleomagnetic laboratory processing of the sam-

ples drilled and oriented in situ in the field included
measurements of the natural remanent magnetization,
stepwise demagnetization by alternating field or thermal
method, and measurements of the magnetic susceptibili-
ty anisotropy.

3. Petrography and volcanology

In this section, we describe and interpret the petro-

graphic and volcanological features of the mapped vol-
canic rocks. Petrographically identical rock types may
occur as different massive rock bodies and volcaniclastic
deposits. In distinguishing genetic rock types, we fo-
cused on the assemblage and relative abundance of phe-
nocrysts and accessory minerals as well as structural
features of the mineral phases. Glass content and phenoc-
ryst versus glass ratio of a rock as well as oxidization
state and alteration of minerals have also been consid-
ered.

Our groups below are based on petrographic types

(Fig. 3) combined with volcanological facies relation-
ships and inferred transport processes. We distinguish be-
tween (1) massive rock types of subvolcanic rocks, dykes
and lava domes/flows, (2) pumice-bearing volcaniclastic
(mostly resedimented) deposits, (3) block-and-ash flow
deposits, and (4) other volcaniclastic (epiclastic) mass-
flow deposits. It is important that all lithic clasts of
groups 2, 3 and 4 are petrographically identical with the
subgroups of massive rock types (1), thus help to reveal
stratigraphic relationships. The most important petro-
graphic properties are summarized in Table 1. All types
are compared to the previously described rocks of the
Börzsöny Mts (Karátson et al. 2000). Locality names are
given in a simplified volcanological map (Fig. 2), a new
compilation of previous data and petrographic and vol-
canological mapping for the present study. The groups
are numbered in agreement with Fig. 2, and for each
group an informal lithostratigraphic name (e.g. Lom Hill
Lava Dome, Holdvilág-árok Tuffaceous Sandstone) is
proposed.

totGm – total groundmass content, totPc – total phenocryst content, opx – orthopyroxene, cpx – clinopyroxene, plag – plagioclase,
opq – opaque phase, totMafic – total mafic content, n.o. – not observed.

Table 1: Textural parameters and phenocryst modes of the massive rock types of the Visegrád Mountains.

547

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

3.1 Massive rocks (subvolcanic and extrusive products)

3.1.1 Dacites/rhyodacites

On the basis of whole-rock chemistry (section 4), the

two subgroups described here belong to the dacite/rhyo-
dacite rock type in spite of the lack of quartz as free
phase.

Garnet-bearing biotite dacite (1a Lom Hill Lava

Dome). This is a light grey or pale greenish grey effusive/
shallow level subvolcanic rock with hyalopilitic to tra-
chytic texture and commonly flow-band structure and
high glass content. The groundmass is generally inhomo-
geneous: brown, glass-rich hypohyaline patches exist in
more transparent, colourless, predominantly hypocrystal-
line matrix, which is relatively rich in microphenocrysts.
In the trachytic textured lava flows the arrangement of
these patches in clusters and their alignment in bands
cause the common macroscopic flow-band structure of
these rocks, well-known for a long time (Zelenka 1960).
The groundmass versus phenocryst ratio is much higher
compared to the andesites (Table 1). The main phenoc-
rysts are plagioclase (andesine to labradorite) and biotite,
while opaque phase (magnetite/titanomagnetite) and al-
mandine garnet occur subordinately. Biotite occurs both
as unaltered entire crystals and as partially opacitized and
resorbed fragments suggesting comagmatic origin for the
former and xenocrystal origin for the latter population.
Garnet is a comagmatic (Harangi et al. 2001) and charac-
teristic accessory phase. Biotite and plagioclase as well
as zircon and apatite are common inclusions in the gar-
net. The latter two minerals generally occur in the
groundmass too, thus they are characteristic accessory
phases of this dacite group, while they are absent or
scarce in the andesites.

In the Visegrád Mts, the garnet-bearing biotite dacites

occur mainly at the S and SW periphery (including the
most voluminous Lom Hill: Fig. 2). High glass content of
the rock, as well as roundish, isolated morphology of
many hills consisting of this dacite type (Fig. 2), and anal-
ogy to the Börzsöny Mts, argue for the existence of small-
to moderate-sized extrusive bodies, such as lava domes.

Pyroxene dacite± garnet (orthopyroxene; 1b Csódi

Hill Laccolith). This is a light grey shallow-level subvol-
canic and effusive rocks with hyalopilitic texture and
rare phenocrysts. The groundmass is hypocrystalline
with a high amount of colourless glass, while the needle-
shaped plagioclase and orthopyroxene microphenocrysts
(up to 50 m), as well as very few opaque microcrysts,
are subordinate (Table 1). Biotite is absent from the
groundmass. The phenocryst assemblage is plagioclase
(andesine to labradorite), hypersthene with minor opaque
phase, and biotite. Unlike in the 1a dacite type, biotite
occurs only as small (up to 1 mm) resorbed and fully
opacitized crystals suggesting xenocrystal origin, while
garnet is less frequent or absent in some localities (e.g.
Peres Hill).

This rock type occurs as eroded lava domes in the S

periphery of the mountains (e.g. Peres Hill), and as a sub-

volcanic body (laccolith) of Csódi Hill in the NE (Koch
1871; Harangi 1999).

3.1.2 Andesites

Most of the volcanic rocks in the study area, ca. 90 %,

are andesites. We distinguish between four andesite sub-
groups that represent the majority of the andesitic rocks
of the mountains:

Biotite amphibole andesite (biotite, hornblende; 2a

Ördögbánya Lava Dome). This is a dark grey or brownish-
grey typically effusive rock type. The texture is hyalopil-
itic with a high amount of glass and high groundmass/
phenocryst ratio (Table 1). The opaque phase is almost
fully absent in the groundmass, being one of the distinc-
tive features of this rock type. The phenocryst assemblage
is plagioclase (andesine to bytownite), amphibole, biotite
and magnetite, while orthopyroxene sporadically occurs
as trace phase. The majority of the sparse hypersthene
crystals (which may be absent) are fragmented and many
crystals show resorption rims indicating disequilibria be-
tween the magma and the crystals, which suggests a xe-
nocrystal origin of the pyroxene, while the biotite is
surely comagmatic.

This rock type occurs in two large bodies (e.g. Ördög-

bánya Quarry). They may represent deeply eroded, dis-
sected large bodi(es), possibly of lava domes, similar to
Pap Hill and Nagy Sas Hill biotite andesites (Karátson et
al. 2000) in the Börzsöny Mts, because we found this
rock type in heterolithic, epiclastic breccias (e.g. Kő
Hill), so it should have originated, at least in part, from
an extrusive rock.

Pyroxene amphibole andesite (orthopyroxene,

oxyamphibole: 2b Keserűs Hill Breccia). The colour of
this rock, changing by locality, depends on the redox
condition. The texture is hyalopilitic, but
microphenocrysts are relatively abundant in the
groundmass (Table 1). In the highly oxidized samples
both the amphibole microphenocryst phase and the
opaque phase (magnetite, secondary hematite) are red in
colour. This alteration was caused by syneffusive
oxidation of the hot lava dome surface in a subaerial
environment. At other localities amphibole is brown and
magnetite is unaltered, so the rock colour is light grey.
The phenocryst assemblage is plagioclase (andesine to
bytownite), oxyamphibole, and magnetite as minor
phase. Orthopyroxene occurs as a minor phase in most
samples, but sometimes it exists as a major phase. The
amphibole/pyroxene ratio is much higher than in the
next group. Biotite is generally absent, but in some sam-
ples it can occur as minor phase (Table 1). The biotite-
free and biotite-bearing varieties can be found at the
same places, thus they may belong to the same group.
Apatite occurs as a rare but constantly present accessory
phase only in this andesite type. There is an enriched,
biotite-bearing subtype of 2b that is a dark red, highly
oxidized pyroxene-amphibole andesite and its main
phenocryst assemblage is similar to the 2b type but
biotite phenocrysts occur as a minor ( < 1.5 %) phase. All

548

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

the mafic minerals are fully opacithized and the pristine
opaque phase is replaced by hematite. A real peculiarity
of this subtype is the high abundance of accessories.
Large (up to 300 m), entire euhedral zircon as well as
apatites that are short prismatic euhedral crystals (up to
500 m) are frequent. In several samples, the amount of
this latter phase approaches the minor phase category
near to 1 % abundance. The majority of these crystals
exist as inclusions in the opacithized amphiboles, but a
lot of them occur as a free phase in the groundmass or as
inclusions in plagioclases.

The 2b rock type occurs as a local lava flow in the

northernmost part of the mountains (vicinity of Hegyes-
tető Hill), and is the dominant constituent of block-and-
ash and debris-flow deposits in the upper volcaniclastic
unit of the N part of the mountains, dispersed all around
Keserűs Hill volcano. At Visegrád Castle Hill it compos-
es the upper monolithic block-and-ash flow breccias,
whereas at Szent Mihály Hill middle W part they can be
found in mono- and heterolithic debris-flow deposits
(Fig. 4, section 4).

Pyroxene amphibole andesite (orthopyroxene, horn-

blende: 2c Szent Mihály Hill Lava Dome). This is a
brown or greyish-brown rock type. Its texture is predomi-
nantly hyalopilitic, but in some lava flow units it contin-
uously changes laterally from hyalopilitic to trachytic
with decreasing glass content. The phenocryst assem-
blage is plagioclase (labradorite to bytownite), horn-
blende and orthopyroxene as major phases, and
magnetite as a minor phase (Table 1). In the majority of
the samples hornblende is the main mafic mineral, but in
a few localities orthopyroxene is predominant. Biotite is
absent. The hornblende is frequently opacitized and hy-
persthene commonly also shows opacite-like alteration.
The opaque phase occurs both as microphenocrysts
( < 20 m) in the groundmass and euhedral phenocrysts
(150—400 m), while intermediate-sized crystals are
lacking. This bimodality of crystal size, which is a char-
acteristic feature of the rock, suggests a two-stage crys-
tallization process (Hibbard 1995).

This rock type constitutes the large bodies of Szent

Mihály Hill and Ágas Hill Lava Domes (massive rock,
lava breccia and minor block-and-ash flows). Also, it is a
subordinate block constituent of block-and-ash flow de-
posits in the N region of the mountains (Rocks of
Vadálló-Kövek, Visegrád Castle Hill) and volcaniclastic
mass-flow deposits toward the periphery (Hideglelős
kereszt and Basaharc Valley in the NW, Hirsch orom Hill,
Rocks of Zsivány-sziklák in the S, Csikóvár Hill, Kő Hill
and Ravine of Vasas-szakadék in the SE etc.). Further-
more, 2c type lithic clasts occur in a stratigraphically
low position in the phreatomagmatic deposits of the ra-
vine of Holdvilág-árok (Fig. 4 section 1), representing
one of the earliest volcanic products of the mountains.

Basaltic andesite (ortho-, clinopyroxene, amphibole)

(2d Dömör-kapu Lava Dome). This is a group of grey or
brownish-grey shallow level subvolcanic rocks, such as
dykes. The texture is generally pilotaxitic. Phenocrysts
are abundant, especially plagioclase; the plagioclase/

whole mafic phase ratio is the highest in the andesite
types (Table 1). The phenocryst assemblage is plagio-
clase (bytownite to anortite), clinopyroxene and ortho-
pyroxene. In addition to magnetite, amphibole can also
be found as a minor phase. The subordinate amphibole is
almost completely altered by opacitization. Remnant
cores, commonly surrounded by a corona, consist of
small hypersthene crystals, which show that both horn-
blende and oxyamphibole are original phases suggesting
inheritance from the preceding andesite types. In some
localities the groundmass is modified by subsequent or
late-stage syneffusive hydrothermal alteration manifest-
ing itself by small patches of spherulitic saponite, tridim-
ite or opal assemblages.

This rock type is petrologically similar to the late-

stage basaltic andesites of the Börzsöny Mts (Karátson et
al. 2000). In the Visegrád Mts it occurs either as medium-
sized subvolcanic bodies (Mátyás Hill Laccolith) to the
deep part of lava domes (Dömör-kapu Lava Dome), or
lava flows (Tövises Hill Lava Flow). The lava flows de-
velop striking platy jointing and are characterized by
uniform, outward dips (170—180/22

), implying an ori-

gin from the Keserűs Hill dome complex. The moderate
dip also points to a location on the middle slope of the
volcano (Fig. 2). (These platy jointed lava flows are typi-
cal in the High Börzsöny dome complex (Karátson et al.
2000) but except for Tövises Hill missing from Keserűs
Hill.) Platy jointed basaltic andesite also occurs on Szent
Mihály Hill lower part, and a strongly brecciated lava or
dyke occurrence on its middle part.

3.2 Pumiceous mostly resedimented volcaniclastic de-
posits

Pumiceous, fine-grained volcaniclastic deposits (PVD)

are exposed in a great number of localities all over the
mountains (Fig. 2). They seldom exhibit evidence of hot
emplacement, such as segregation pipes (e.g. ravine of
Holdvilág-árok upper part, quarry: Fig. 4, layer 1/F) or
unbroken crystals in fine ash (fallout deposits at Rocks
of Zsivány-sziklák: Kósik 2005). In contrast, most depos-
its contain abundant, commonly subrounded, cm- (occa-
sionally dm-) sized pumice clasts, typically set in a
volcanogenic sand- to claystone matrix occasionally
with claystone- and quartzite pebbles (Öregvíz Stream,
the Ravines of Salabasina-, Holdvilág-árok and Rám-sza-
kadék, Vaskapu and Ráró Hills, Szakó Hill E, Kő Hill
etc.: Fig. 4). In some PVD, subangular to subrounded
andesite lithic clasts are also present (e.g. Rocks of
Zsivány sziklák, Ravines of Holdvilág-árok and Vasas-
szakadék, Öregvíz and Sztelin Streams, Őr Hill, etc.:
Fig. 4).

The occasional high pumice content implies a time-

space relationship to the original pyroclastic processes.
However, (a) the common normal grading of lithic and
reverse grading of pumice clasts, (b) the existence of al-
tered and/or subrounded pumice clasts and (c) the fre-
quent non-volcanic xenolith content points to epiclastic
reworking. Erosive channels, alternation of pumice-rich

549

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

lenses with pumice poor laminae, typically <1 m-thick
flow units, as well as frequent bedding and grading, im-
ply deposition from volcaniclastic density currents. Re-
working by subaqueous to subaerial debris-flows seems
to have been the common range of transport processes
(cf. Konečný & Lexa 1994; Karátson & Németh 2001;
also see point 3.4).

In stratigraphical order, the following types can be dis-

tinguished as informal lithostratigraphic units.

3.2.1 Csádri Valley Sandstone

At the base of some valleys there is evidence of sub-

aquaeous/submarine deposition of fine-grained volcani-
clastic mass-flows (e.g. Ostrea and Aequipecten seniensis
fragments embedded in pumiceous and andesite lithic-
bearing volcanogenic sandstone), especially in the E pe-
riphery of the mountains (e.g. Csádri Stream at Duna-
bogdány). For further details of these deposits, see
section 7 on paleontology.

3.2.2 Holdvilág-árok Tuff and Tuffaceous Sandstone

In the lower position of other localities, tuffs, lapilli

tuffs and tuffaceous sandstone ( ± pumice) were deposited
in water-saturated environment (marked by accretionary
lapilli, fine lamination, well-developed, contrasting
grading of lithic and rounded pumice clasts: ravines of
Holdvilág-árok and Vasas-szakadék, Kő Hill, Sztelin
Stream, Visegrád Fekete Hill base (“Panorama Road”),
Kisvillám Hill, ravine of Rám-szakadék lower part). The
earliest volcanic eruptions of the mountains, as revealed
in the lowermost units of the ravine of Holdvilág-árok
(Fig. 4/1), produced phreatomagmatic deposits (i.e. with
accretionary lapilli) and reworked pumiceous epiclastic
rocks, directly overlying the basal marine sediments (see
section 6 on Paleomagnetic data). There, the bulk chem-
istry of pumices is andesitic (Bendő et al. 2001). Their
lithic clasts and the groundmass contain hornblende and
orthopyroxene only, while garnet, biotite and oxyam-
phibole are absent, suggesting a 2c type andesite. Accre-
tionary lapilli-bearing andesite tuff overlain by dacitic
pumiceous lapilli tuff is also exposed at Sztelin Stream
(Fig. 4). Korpás et al. (1967) described “andesitic tuffs”
in basal position from Malom Valley, the NE part of the
mountains, too. There, at Hirsch orom Hill, we have
found 2c type andesite lithic clasts in a pumiceous mass-
flow unit. In many localities, the shallow-water eruptions
gave place to emergent activity, as charcoal or treetrunk
remnants show (Kisvillám Hill – Hegedűs 1953; Rocks
of Zsivány-sziklák – Kósik 2005; Dobogókő, Ecset and
Őr Hills) as well as leaf imprints (Kisvillám Hill –
Hegedűs 1953; Dobogókő Hill – Zelenka 1960). This
implies a gradual shift from shallow submarine to sub-
aerial deposition (and/or a nearby terrestrial environ-
ment). Rhyolitic-dacitic composition has been pointed
out from only the ravine of Holdvilág-árok (Bendő et al.
2001) and Öregvíz Stream (i.e. SE part of the mountains)
where PVD seem to be related to local lava dome activity

(e.g. Zelenka 1960; Höfer 2003). (Without chemical anal-
ysis, Koch (1877) and Zelenka (1960) also mention “dac-
ite tuffs” and “tuffites”.) In the ravine of Holdvilág-árok,
mixed, pumice-poor debris-flow deposits that contain lith-
ic clasts both with andesitic (2c type) and dacitic (1a type)
composition as well as pumices both with andesitic (horn-
blende- and hypersthene-bearing) and dacitic/rhyolitic
composition (biotite-bearing) appear in a low stratigraphic
position. They are overlain by pumice-rich reworked epi-
clastics that contain biotite-bearing pumice clasts with
rhyolitic composition. A hot emplaced primary ignimbrite
unit that is in a higher stratigraphic position contains
andesitic pumice clasts and its mineral assemblage is iden-
tical with the 2c pyroxene-amphibole andesite, suggesting
a concordant magmatic origin. Oxyamphibole fragments
as well as andesite lithoclasts of 2b andesite type appear
only in the overlying pumice-poor reworked deposits,
while augite and 2d basaltic andesite fragments occur in
deposits of the uppermost level.

3.2.3 Rám Hill Pumiceous Sandstone

A characteristic PVD occurs in the SW and SE fore-

ground of Keserűs Hill volcano (e.g. valley heads of Rám
Hill to upper level of Ravine of Rám-szakadék; Öregvíz
Stream) as well as further to the W (valleys of Maróti
Hills). This epiclastic rock is full of mm- to cm-sized, 2c
type pumice fragments. It also contains claystone,
quartzite pebbles and rarely mollusc fragments, and may
also have been resedimented directly from explosive
eruptions in a shallow water environment.

For all these types of PVD, brecciated lava domes as

well as small explosive eruptions can be envisaged as
major sources, similar to those inferred in the Burda
(Helemba) Mts (Konečný & Lexa 1994) and Börzsöny
Mts (Karátson & Németh 2001). Garnet- and dacitic li-
thoclast-bearing PVD may have been originated from the
neighbouring dacite lava domes (e.g. Höfer 2003). Some
types of PVD, most typically the Rám Hill Pumiceous
Sandstone, may have been originated from Keserűs Hill
Lava Dome itself, as indicated by its distribution and its
2c type andesite, which is found as dykes in the source
area. The described localities occur both inside and out-
side the mountains suggesting that transport processes
and paleogeography were generally the same all over the
mountains. This conclusion is in accordance with scat-
tered, small- or medium-sized volcanic centres.

3.3 Block-and-ash flow deposits

Mostly in the upper levels of the mountains (typically

around Keserűs Hill volcano: ca. 35 km

), a great number

of outcrops reveal coarse-grained pyroclastic breccias
(clast size up to 5 m) with monolithological composition
(Keserűs Hill Breccia: Fig. 4). Petrographically, the main
( > 95 %) block constituent is the 2b pyroxene amphibole
andesite, but there is a unique, zircon and apatite rich bi-
otite-bearing subtype that is a minor ( < 5 %) but spatially
widespread constituent. In some localities the 2c type

550

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

Fig. 4. Selected stratigraphic logs of the dacitic-andesitic volcanism of the Visegrád Mts. For localities, see Fig. 2. 1 – Ravine of Hold-
világ-árok (No. 1 in Fig. 2, Table 3): A – Ostrea-bearing siltstone, B – Andesitic phreatomagmatic units with accretionary lapilli, C –
Garnet-bearing rhyodacitic phreatomagmatic units, D – Andesitic block-and-ash flow, E – Andesitic pumice-bearing hyperconcentrated flood-
flow and debris-flow, F – Andesitic ignimbrite, G – Andesitic fall and small-volume pyroclastic-flow, H – Andesitic debris flow, J –
Rhyodacitic block-and-ash flow units inferred to be related to coherent rhyodacite bodies, K – Andesitic debris-flow units of block-and-
ash flow origin. 2 – Sztelin Stream: A – Submarine (?) siltstone with pumice fragments, B – Siltstone, C, E – Bedded siltstone, D – Pum-
ice-rich tuffaceous sandstone, H – Subaerial (?) dacitic volcaniclastic mass-flow unit with pumice-rich lens at base, I – Pumice-rich lapillis-
tone, J – Pumice-bearing bedded, cross-bedded lapilli tuff with pumice concentration lenses and zones, K – Siltstone (hyperconcentrated
flood-flow) with pumice fragments, L – Pumice- and andesitic lithic-rich lapilli tuff, M – Coarse volcanogenic sand (fluvial deposit), N –
Andesitic block-and-ash flow with pumice fragments, O – Volcaniclastic debris flow. 3 – Kő Hill: A – Andesitic debris flow, B – Volcan-
ogenic siltstone, C – Debris flow with reverse grading, D – Andesitic lapilli-bearing fine breccia, E – Volcanogenic siltstone, F –
Pumiceous lapilli tuff/tuffaceous sandstone, G—G1 – Volcanogenic sandstone (hyperconcentrated flood-flow), H—I – Andesitic debris-flow
units with normal and reverse grading and channel fills, J – Silty sandstone (hyperconcentrated flood-flow), K – Andesitic debris flows
with normal grading and lithic concentration zones. 4 – Szent Mihály Hill southwestern part (Nos. 34, 14, 35): A – Strongly fractured
andesite subvolcanic body, B – Medium-grained breccia (debris flow), C – Densely packed ungraded debris flow, D – Reversely graded
debris flow, E – Tuffaceous sandstone, F – Heterolithic, very coarse-grained breccia (small-volume debris avalanche or big debris-flow),
F1 – Monolithologic breccia, G – laminated siltstone (hyperconcentrated streamflow), H – Slightly bedded fine-grained breccia, J –
Reversely graded debris flow, K – Coarse-grained monolithological debris flow, L – Strongly fractured andesite lava breccia (summit
lava dome), M – Basaltic andesite lava breccia (dyke/sill?). 5 – Rocks of Vadálló-kövek (No. 20). 6 – Rocks of Thirring-sziklák (No.
24): A – Andesitic debris flow with reverse grading, B – Fluvial streamflow/debris flow, C – Andesitic debris flow with minor reverse
grading, D – Densely packed, reversely graded debris flow, E – Bedded hyperconcentrated flow, F – Normal graded debris flow, G –
Debris-flow with double grading, H – Slightly reversely graded debris flow, I – Ungraded debris flow. 7 – Rock of Zsivány-sziklák: A –
Erosion channels in pumiceous, lithic-rich debris flow, B – Reversely graded pumiceous volcaniclastic mass-flow, C – Lithoclast-
bearing, slightly pumiceous diluted debris-flow, D – Channel fills of lithic- and pumice-rich debris flow. E – Slightly pumiceous debris flow, F,
H – hyperconcentrated flood-flow, G, I, J – Andesitic block-and-ash flow-related debris flow sequence. Continued on next page

551

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

also appears, but it is scarce ( < 1 %) and presumably of
accidental (picked-up) origin. Poor sorting and stratifica-
tion, no or minor grading, frequent prismatic jointing
and porous texture of clasts are common at the Keserűs
Hill—Öreg Pap Hill ridges. Scarcely, small (centimetric)
pumice fragments can also be found. These features are
consistent with a proximal facies of subaerial block-and-
ash flow deposits (BFD). As for the source area of BFD,
observations and analogies indicate a single, cone-
shaped lava dome group (i.e. the Keserűs Hill edifice):
(1) radial pathways inferred from dip data (Karátson et al.
2006, also see Fig. 2), (2) monolithological composition,
(3) systematic facies changes from proximal BFD to pet-
rographically identical debris-flow deposits (e.g.
Dobogókő Hill 3 km and Csikóvár Hill 7 km to the S, Ecset
Hill 5—6 km to the W, Borjúfő Hill 5 km to the E, dis-
tances with respect to the Keserűs Hill volcano geometri-
cal centre), and (4) geomorphic, volumetric analogies to
worldwide examples (i.e. Unzen, Mont Pelée; see calcu-
lations in Karátson et al. 2006). This way, the inferred
source area should have been around Ágas Hill, which it-
self, however, is a post-BFD feature (i.e. another, smaller
lava dome of 2c type andesite) that formed subsequent to
the proposed sector collapse. On the other hand, the
close proximity of BFD to the centre (e.g. Rocks of
Vadálló-kövek: 2.5 km) does not mean that the upper
part of the Keserűs Hill volcanic cone has been pre-
served. During the 15 Myr-long degradation, the original
horseshoe-caldera of the volcano has been strongly erod-
ed, its rim lowered and retreated (Karátson et al. 2006),
which is proven by the fact that these proximal BFD are
interbedded with undulating, hyperconcentrated to fluvi-
al streamflow deposits with only 5 to 15º dip, emplaced
originally on the lower flanks (Fig. 4; also see point
3.4.3).

In the BFD, the described colour change of blocks (2b

type) fits the inferred lava dome origin: it represents the
contemporaneous oxidation of the fracturing lava domes.
This feature, coupled with the pumice content as well as
the porous andesite clasts, implies explosive dome col-
lapses rather than simple gravitational ones. It is note-
worthy that the High Börzsöny BFD (Karátson 1995) do
not show these features, possibly due to the dominant
gravitational dome collapses inferred.

The above mentioned facies changes toward debris

flows show the extent of the area where the coarse-
grained part of the original BFD were still deposited and
preserved. The debris-flow counterparts are better sorted
and graded and are mostly composed of thinner flow
units (e.g. Dobogó-kő Hill: see Fig. 4). The existence of

the monolithologic block-and-ash-flow—debris-flow con-
tinuum beyond the previously proposed “outer caldera”
margin (i.e. Dobogókő or Urak-asztala Hills etc.: Chol-
noky 1937; Korpás (Ed.) 1998) excludes the presence of
an outer “caldera rim” (see earlier).

3.4 Other volcaniclastic mass-flow deposits

3.4.1 Debris-flow deposits (DFD)

In addition to the diluted, monolithological debris-

flows of block-and-ash flow origin, heterolithic DFD
crop out at the most distal localities (relative to Keserűs
Hill volcano), such as Ráró and Ecset Hills in the W,
Csikóvár Hills in the S, Nyerges and Vörös-kő Hills in
the E, Szent Mihály Hill lower-central part in the N.
There, mostly 2c type andesite blocks and lapilli are
mixed with other andesite types (2b > 2d > 2a) and subor-
dinately dacite clasts to various extents, suggesting local
depocenter areas at the contemporary erosion base level.
Lack of widespread conglomerates (that occur subordi-
nately in the ravines of Rám-szakadék or Holdvilág-
árok) implies short transport distances as well as a rapid
change from shallow-water to emergent activity. The
DFD show thinner and more developed flow units than the
diluted block-and-ash flow deposits, and are frequently in-
terbedded with hyperconcentrated to normal streamflow
deposits. Sometimes minor pumice content makes a transi-
tion towards pumiceous volcaniclastic deposits.

3.4.2 Debris-avalanche deposits (DAD)

The best exposed area of coarse-grained volcaniclastic

breccias is the very steep Szent Mihály Hill (exposed by
the Late Pleistocene downcut of the Danube Bend:
Karátson 2001; Karátson et al. 2006). Especially at its W
lower to central part, mono- and heterolithic breccias of
2c—b—d type andesites with block size up to 6—8 m crop
out (e.g. Dobozi-orom Debris Avalanche). The presence
of large/oversized and cracked blocks, sometimes with
jigsaw fit, as well as strongly undulated basal layer point
to small-scale debris avalanche or huge debris-flow trans-
port mechanisms. Given the general northward dip of
beds, we postulate that most of these breccias originated
from the Keserűs Hill volcano, ca. 4 km southward, that
is also characterized by 2b—c—d type andesites. Another
DAD area is the Visegrád Castle Hill upper part (Karátson
2001; Bendő 2002). There, 2a type brecciated lava is
overlain by typical 2b type block-and-ash-flow breccias
that show cracked blocks and jigsaw fit, and the matrix is

8 – Ravine of Rám-szakadék: K – Fine-grained volcaniclastic mass-flow, L – Andesitic debris flow—fluvial streamflow sequence
(L1 – rounded, medium-sized lithics, L2 – wedging of fine-grained bed, L3 – semirounded cobbles in lenses and zones), M – Stratified
andesitic debris flow, N – Ungraded andesitic debris flow, O – Hyperconcentrated flood-flow, P – Andesitic debris-flow sequence
(P1 – Ungraded bed, P2 – reverse grading with oversized clasts on top), Q – Andesitic debris flow with undulating (erosive) base,
R – Hyperconcentrated flood-flow, S – Reversely graded heterolithic andesitic debris flow, T – Andesitic debris flow from block-
and-ash flow, U – Fine lithoclasts in andesitic volcaniclastic mass-flow, V – Monotonous ungraded andesitic debris-flow, Y – Lithic-
rich channel fills in pumiceous volcaniclastic mass-flow.

552

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

characterized by reddish, fine-grained plastic deforma-
tion, typical of debris avalanches. The positive, concen-
tric morphology of the hill might be due to the strongly
resistant coarse breccia cover.

3.4.3 Hyperconcentrated flood-flow (HFD) and fluvial

streamflow (FSD) deposits

HFD and FSD occur dominantly in the peripheral areas

of the mountains (Szent Mihály Hill S part lower se-
quences, Vaskapu Hill, Ecset and Ráró Hill, Rocks of

previously mentioned red and light andesite clasts (of 2b)
of the same block-and-ash flow deposits. Comparing ma-
jor element data of the Visegrád andesites with the
nearby Börzsöny rocks (Karátson et al. 2000), slight dif-
ferences can be observed: the Visegrád andesites contain
a little bit less K

O at a given SiO

content.

The N-MORB normalized trace element patterns

(Fig. 5b) of the basaltic andesites (2d) and the more silicic
andesites have very similar character and show typical
subduction-related features: enrichment of the large-ion
lithophile elements (e.g. Rb, Ba, K), negative Nb-anomaly

1981). The shaded field indicates the Miocene volcanic rocks of the Börzsöny Mts. b –
N-MORB (Pearce & Parkinson 1993) normalized trace element patterns of representa-
tive samples of different groups of the Miocene volcanic rocks of the Visegrád Mts.
Symbols are explained in Fig. 5a. c – Th/Y versus Nb/Y diagram for the Miocene
volcanic rocks of the Visegrád Mts. SZE = subduction zone enrichment; FC = fractional
crystallization; AFC = assimilation and fractional crystallization; E-MORB = enriched
Mid-Ocean Ridge Basalts; OIB = ocean island basalts.

Zsivány sziklák, Öregvíz and Sztelin
Streams etc.), interbedded in particular
with fine-grained pumiceous volcani-
clastic and debris-flow deposits. Unique-
ly, they also crop out in the Rocks of
Vadálló Kövek (Keserűs Hill: Fig. 4),
separating two voluminuous block-and-
ash flow units. Their appearance in basal
strata (e.g. ravine of Holdvilág-árok, and
Őr, Vaskapu and Vöröskő Hills) indicate
that the earliest shallow-water environ-
ment should have rapidly been replaced
by terrestrial environment on the emer-
gent islands. The abundance (wide-
spread preservation) of such deposits,
along with various debris-flow deposits,
implies long, quiescent inter-eruptive
periods, and a relatively low and mosaic-
like relief that did not enable reworking
processes to deposit their load far away.
The emergent archipelago (under sub-
tropical climate), characterized by sub-
aerial debris-flows—hyperconcentrated
flood-flow—fluvial streamflow processes,
was also pointed out in the time-space
related Börzsöny Mts (Karátson &
Németh 2001).

4. Geochemistry

In general, the Miocene volcanic rocks

of the Visegrád Mts are mostly andesites
(Table 2), falling right into the boundary
between the medium-K and high-K calc-
alkaline series in the SiO

versus K

O dia-

gram (Fig. 5a). Basaltic andesites (2d
type) are among the latest rocks and they
occur subordinately. The early-stage vol-
canic rocks have a bimodal character:
they are high-K dacites to rhyodacites (1a
and 1b) with subordinate andesites (2c:
i.e. initial phreatomagmatic activity). The
bulk-rock geochemistry of 2c—b—a andes-
ite types overlaps in places, but shows a
general trend towards increasing SiO

and

O content (Fig. 5a). No compositional

differences can be observed between the

Fig. 5. Geochemical
diagrams of the Viseg-
rád Mts. a – Classifi-
cation of the volcanic
rocks in the SiO

ver-

sus K

O diagram (Gill

553

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

and positive Pb-anomaly. The garnet-bearing rhyodacites
(1a—b) have a similar trace element pattern with slightly
higher incompatible trace element abundances, but they
show a marked depletion in Y and heavy rare earth ele-
ments. A similar feature was also observed in the garnet-
bearing dacites of the Börzsöny Mts (Karátson et al. 2000;
Harangi et al. 2001). The garnet-free andesites of the Vi-
segrád Mts have fairly similar trace element patterns to the
Börzsöny andesites, although the latter ones have charac-
teristically higher Ba concentration.

The basaltic andesite to andesite suite of the Visegrád

Mts forms a linear trend in the SiO

versus K

O diagram,

suggesting a genetic relationship. The cogenetic rela-
tionship is also confirmed by the strong linear correla-
tion between the highly incompatible trace elements,
such as La, Th, Nb, Rb, Ba and Pb. The rhyodacites have
similar incompatible trace element ratios to the andes-

ites, too, implying that they could have derived from a
similar source region. In the Nb/Y versus Th/Y ratio dia-
gram (Fig. 5c), they form a linear trend above the mantle
array. It suggests subduction-related enrichment fol-
lowed by fractional crystallization or assimilation com-
bined by fractional crystallization. Harangi et al. (2001)
proposed that contamination by lower crustal metasedi-
mentary rocks could explain the occurrence of the gar-
net-bearing volcanic rocks. The low Y and heavy REE
content indicate either garnet or hornblende fraction-
ation during the early stage of magma evolution.

5. K/Ar geochronology

In Table 3, along with all previous dates (Balogh 1977—

1979), analytical results and calculated K/Ar ages are given.

Table 2: Main and trace element geochemistry of selected rock types of the Visegrád Mountains.

554

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

The garnet-bearing dacitic rocks have yielded ca.

16 Ma ages (15.9 Ma average age of biotite separates).
The same ages ( > 16 Ma on whole rock and an amphibole
separate) were also obtained from the stratigraphically
related earliest andesitic volcaniclastics, from the ravine
of Holdvilág-árok. Thus, the 16 Ma age may reflect the
beginning of volcanic activity in the Visegrád Mts. Ac-
cording to stratigraphic constraints, the volcanism con-
tinued to develop after a short quiescence, and on the
basis of all available K/Ar data the peak of the volcanic
activity occurred ca. 15.5—15 Ma ago. It should be noted
that, in some cases, andesite K/Ar ages older than 16 Ma
might be the consequence of the presence of excess ar-
gon. Especially andesitic rocks possessing phenochrysts
of olivine, pyroxene, hornblende or plagioclase may
give anomalously old apparent ages, owing to the incor-
poration of excess argon from the environment, as the
phenochrysts crystallized in the magma prior to its erup-
tion (see K/Ar ages obtained on amphibole separates).

A small number of younger ages obtained on whole-

rock samples (i.e. 13—14 Ma) can be related to minor or
secondary event(s). However, at least those K/Ar ages de-
termined on volcanic glasses make it likely that the dates
younger than about 14.5 Ma can rather be considered
“apparent ages” (also see the Discussion).

6. Paleomagnetic data

A great number of paleomagnetic results, reported in

Table 3, were obtained in the 1970s (Balla & Márton-
Szalay 1979), while another set was produced recently.
Since our results are published for the first time, new data
are shown with statistical parameters k and

(Fisher

1953). The scatter of the individual directions within a
sample group (consisting of 4—9 independently oriented
cores) is expressed by k, while

gives the radius of the

confidence circle (which for the previously published lo-
calities varied between 5 and 17).

The newly sampled volcanic rocks are of good quality

and yielded statistically excellent results to infer reliable
paleomagnetic directions. The single claystone locality
from the ravine of Holdvilág-árok (No. 1 in Table 3) is of
lower quality, yet its counterclockwise (CCW) rotation
(with reverse polarity) obtained is important for the gen-
eral interpretation of the paleomagnetic data. The clay-
stone is overlain by initial, andesitic biotite-bearing
phreatomagmatic deposits, whose declinations are also
suggestive of CCW rotation (probably with a successive,
normal polarity). The same characteristics typify the ini-
tial pumiceous deposit of Öregvíz Stream also in the SE
part of the mountains (No. 2). All these biotite-bearing
pyro- and volcaniclastic rocks (occasionally with garnet)
are among the oldest volcanic products in the area. In the
Börzsöny Mts, the basal garnet-bearing dacitic succes-
sions also exhibit CCW rotation (with normal polarity).
Therefore it stands to reason to seek correlation between
the Börzsöny and Visegrád Mts, starting the volcanic
evolution with the CCW rotated normal polarity paleo-

magnetic zone (marked with circle) and proceeding as
Table 3 shows.

A striking feature of the paleomagnetic results from the

Visegrád Mts is the dominance of reversed polarity. In
fact, with the exceptions of the above-mentioned two
normal polarity rotated sites of the earliest successions,
the other positive inclinations, coupled with uncertain
declinations, can be regarded as “anomalous” paleomag-
netic directions (Csódi Hill second site, Szamár Hill and
Hideglelős-kereszt quarries). This way, subsequent to the
oldest volcanic stage with normal polarity, the volcanic
activity should have been most intense during a reversed
polarity zone. As for Csódi Hill, additional results of M.
Lantos (unpublished) complete the picture about a cool-
ing subvolcanic body (e.g. Bendő & Korpás 2005): the
fast cooling edge shows reverse polarity, while the slow-
cooling near-vent part may have been solidified during a
transitional period to normal polarity.

At the same time, in agreement with our conclusion for

the Börzsöny (Karátson et al. 2000), this reversed zone is
characterized by a marked change from counterclockwise
(CCW) rotated declinations to non-rotated ones (with re-
spect to the stable European reference declination). In
Table 3, these two groups are marked with squares and
rhombs. By analogy with the Börzsöny, the rotation peri-
od can be placed within the long reversed polarity zone
between 16 and 15 Ma (see point 8.1 in Discussion).

The paleomagnetic data set of Table 3 does not con-

tain evidence for the continuation of the volcanic activi-
ty in the Visegrád Mts less than 15 Ma, although we note
that the basaltic andesite rocks, among which are the
youngest products from stratigraphic constraints, have
not been measured with one exception.

The magnetic anisotropy of some sites (i.e. in the ra-

vine of Holdvilág-árok) shows that the volcanic rocks
were deposited concordantly with the underlying, large-
ly horizontal sedimentary deposits. This implies that no
large-scale tilting has occurred in the mountains. This
conclusion is in accordance with the fact that the mag-
netic foliation planes of the most widespread and mea-
sured 1st stage tuffs in the Börzsöny Mts are also
near-horizontal.

7. Paleontological constraints

In the Visegrád Mts, fossiliferous sedimentary deposits

can be found in smaller areas than in the Börzsöny Mts.
However, there are a number of important key localities
mainly in the E and N periphery, representing (1) the ini-
tial volcanic eruptions, (2) the gap in the volcanic activi-
ty following the dacitic eruptions, and (3) the period just
after the termination of volcanism.

(1) A fossil-rich assemblage similar to the Kismaros

Tuff in the Börzsöny Mts (Báldi & Kókay 1970) is not
known in the Visegrád Mts from the beginning of the
volcanism. However, some old findings along the E pe-
riphery (close to the present Danube: e.g. Öregvíz
Stream, Nagy Villám Hill etc.) reported by Koch (1877)

558

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

and Wein (1939), as well as those of the W mountain mar-
gin (Vaskapu Hill at Esztergom: Peters 1869) imply shal-
low-marine environments with signs of volcanic activity.
In our field work, one of the classical localities of Koch
has been reambulated in the Csádri Stream at Duna-
bogdány. There, a Chlamys gigas-bearing Lower Miocene
(Eggenburgian) sand- and claystone (Méhes 1941; Bohn-
Havas & Korecz-Laky 1980) is overlain by a pumiceous,
andesite lithic clast- and quartz pebble-bearing mass-flow
deposit with Aequipecten seniensis (previously Chlamys
scabrella) fragments. This species is known from the Bur-
digalian to the Pliocene in the Mediterranean region. It
also has a long range in the Central Paratethys, from the
Ottnangian to the Badenian, which at the same time con-
firms the post-Eggenburgian beginning of volcanic activi-
ty. The described fossil-bearing deposit is directly
overlain by pumiceous as well as lithic-rich, fossil-free
volcaniclastic layers.

(2) The end of the dacitic volcanism is indicated by a

short sedimentary episode, which can be traced on the S
slopes of Tornyos Hill at Pilisszentkereszt (Zelenka 1960).
The existence of the greyish-white conglomerate contain-
ing quartzite, metamorphic and volcanic (dacite and
andesite) pebbles was confirmed recently by the present
authors. The rock contains lots of fossils, but original shell
materials can be identified for only Crassostrea, Balanus
and tube worm specimens, while the other mollusc species
show external and/or internal moulds. The composition of
the fauna is similar to that described by Zelenka (1960),
with the most frequent fossils Crassostrea, Turritella, Bal-
anus and Cerithium. Besides these taxa, our preliminary
fauna list contains Barbatia, Natica, Cardium, Cypraea,
venerid bivalves, trochid gastropods and tube worms. The
Crassostrea and Balanus shells always show strong erosion
and rounding. The quantity and roundness of pebbles, the
significant erosion of shells, the composition of the fauna,
as well as the presence of herbivorous gastropods unequivo-
cally refer to shallow, near-shore, well-agitated normal ma-
rine environment. The very poor preservation of the fossils,
however, generally makes it difficult to identify the fauna at
species level. The composition of the fauna is largely simi-
lar to the mollusc assemblages known from Middle Mi-
ocene Leitha Limestone localities.

(3) The first fossiliferous sediments deposited after the

termination of volcanism have been recovered in the N
part of the Visegrád Mts (Fekete Hill at Visegrád). Scholz
(1970) identified 9 coral species from the outcropping
Leitha Limestone, which unambiguously indicate a shal-
low, (sub)tropical, normal marine environment. Müller
(1984) described several decapod species there, and referred
to the locality as Early Badenian in age. Recently, the bryo-
zoan fauna of several Hungarian Badenian localities, in-
cluding that of Fekete Hill, has been investigated by
Moissette et al. (2006, 2007). The most frequent bryozoan
morphology types are the encrusting memraniporiform col-
onies on the surfaces of the coral colonies and calcareous
algae, and small roundish celleporiform colonies. Besides
these forms, several cellariiform and reteporiform colonies
and only a few vinculariform and adeoniform ones have

been found in the surrounding sediments. Such a composi-
tion of the colony morphologies refers to a very shallow
marine environment (0—20 m water depth).

8. Discussion: stratigraphy and volcanic evolution

with respect to the Börzsöny Mountains

8.1 Volcanic stratigraphy

1st volcanic stage. Paleontological and volcanological

evidence suggests that just before the volcanic activity
the main area of the Visegrád Mts was characterized by a
shallow subaquaeous environment (e.g. Csádri Valley
Sandstone) which rapidly became subaerial. The first vo-
luminous volcanic products, distributed at the S and SE
margin, are garnet-bearing volcaniclastic and massive
rocks of dacitic-rhyodacitic composition. These are pre-
ceded by small-volume 2c type andesitic pyroclastic (in
some cases phreatomagmatic) deposits both in the S-SE
(ravine of Holdvilág-árok, Sztelin Stream) and the N (vi-
cinity of Malom Valley). These deposits are collectively
called the Holdvilág-árok Tuff and Tuffaceous Sandstone.
The volcanic centres may have been small-scale edifices,
preferably lava domes (e.g. Lom Hill Lava Dome, biotite
dacite) and subvolcanic bodies (e.g. Csódi Hill Laccolith,
pyroxene dacite), as suggested by identical lithologies of
massive and volcaniclastic rocks in some places. Due to
volcanological constraints (e.g. no or minor ignimbrites,
small pumice sizes, mostly reworked pumiceous deposits,
no recognized caldera morphology etc.), large explosive
eruptions and related calderas can be excluded.

In the Börzsöny Mts, a great number of garnet-bearing

dacitic-rhyolitic volcaniclastic as well as massive rocks
have been determined older than 16 Ma by combined K/Ar
geochronology and magnetostratigraphy (Karátson et al.
2000). These rocks have normal polarity with a typical
CCW rotated declination. Two localities of the initial vol-
caniclastic rocks of the Visegrád Mts show the same pale-
omagnetic features. Of them, the ravine of Holdvilág-árok
seems to be highly reliable for (a) its three successive pale-
omagnetic stages pointed out from the underlying sedi-
mentary deposits through the volcanic pile; (b) within the
analytical errors, its 16.0 Ma K/Ar age obtained fits the
proposed 1st stage (16.0—16.5 Ma) and early 2nd stage
( < 16 Ma) activity of the Börzsöny Mts. The average age
(15.9 Ma) of the biotite fractions of all Visegrád garnet-
bearing dacite samples (see Table 3) is also in accordance.
On the contrary, measurements on massive garnet-bearing
dacites are subordinate, and they show different paleo-
magnetic characteristics (most fall into the 2nd stage with
CCW declination but with reverse polarity: 1a type dacite
of Szentlélek Hill and 1b type dacite of Csódi and Peres
Hills). The emplacement of the Csódi Hill Laccolith be-
fore or even after the rotation clearly shows that the 1b
type dacite may have been younger than the most volumi-
nous 1a type dacites. One of the Börzsöny dacites, the
Nagy Pogány Hill Lava Dome, also shows the 2nd stage
(CCW declination, reverse polarity) and, similarly to the

559

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

Visegrád 1b dacites, garnet is absent or accidental. Al-
though these sites should be investigated in more detail,
we can conclude that (1) the initial (biotite- and garnet-
bearing andesitic and dacitic) explosive period started si-
multaneously with the well-defined 1st paleomagnetic
stage of the Börzsöny (i.e. 16 Ma), and (2) at least part of
the related dacitic extrusive activity in the Visegrád side
(i.e. the 1b type dacites) seems to have lasted longer
( 16 Ma), during the 2nd paleomagnetic stage with re-
verse polarity. Toward the end of the dacitic volcanism, a
quiescent period may have taken place at least in the S, as
proven by interbedded submarine deposits and fossils at
Tornyos Hill.

2nd volcanic stage. A number of the Visegrád paleo-

magnetic sites (12) shows CCW declination with reverse
polarity. The rocks of this group are various andesite types
as well as pyroclastic and epiclastic breccias (e.g. Ördög-
bánya Lava Dome, Rám Hill Pumiceous Sandstone). The
rocks are garnet-free. The average of the K/Ar ages (10 dat-
ings) is 15.57 Ma (all datings) and 15.48 Ma (without am-
phibole and glass ages), that is, ca. 15.5 Ma. However, due
to the limited number and partly the earlier datings (with
> 1 Myr error), the real geological age should be consid-
ered with caution.

3rd volcanic stage. The next group comprises sites with

reverse polarity and no rotation (13 measurements). The
much bigger number of K/Ar datings (36, of which 29 are
new measurements) makes the average age reliable:
15.06 Ma (all datings), 14.90 Ma (without 6 amphibole
and glass ages), that is, ca. 15.0 Ma. The 2b type andesites
(Keserűs Hill Breccia, Hegyes-tető Hill Lava Flow), most
of the 2c type andesites (Szent Mihály Hill Lava Dome,
Ágas Hill Lava Dome), and most of the 2d type basaltic
andesites (e.g. Tövises Hill Lava Flow) were formed dur-
ing this paleomagnetic stage.

In the Börzsöny Mts, rocks belonging to the latter two

above-mentioned paleomagnetic stages are also widely
distributed. They were interpreted as successive stages
representing an intense volcanic period, during which a
30º CCW rotation occurred (Karátson et al. 2000). The ro-
tation was part of a general Middle Miocene rotation peri-
od in North Hungary (Márton & Márton 1995). The exact
timing of the rotation movement could not be determined
in the Börzsöny Mts, but on the basis of K/Ar ages and
magnetostratigraphic considerations, the 16.0—14.5 Ma
interval for the total duration of the two stages was pro-
posed (Karátson et al. 2000). Now, in the light of our re-
sults in the Visegrád Mts, average K/Ar ages of rocks
emplaced before and after the rotation constrain the rota-
tion between 15.0—15.5 ( ± 0.5) Ma. Moreover, given the
well-defined emplacement of the Keserűs-Hill breccias
(15.3 Ma) that do not have a westerly declination, the ro-
tation movement should have occurred between 15.5—
15.3 ( ± 0.5) Ma.

Apart from massive rocks, there are pumiceous epiclas-

tic and rarely pyroclastic deposits that belong to the 2nd
and 3rd volcanic (paleomagnetic) stage. In contrast, pumi-
ceous deposits are missing from the 2nd and 3rd stage in
the Börzsöny Mts (except a single, uncertain locality of

Magyarkút in the SE periphery). This difference indicates
a longer-lasting explosive activity in the Visegrád Mts,
which is in accordance with the explosive lava dome char-
acter of the Keserűs Hill volcano.

The age of the Keserűs Hill Lava Dome complex, the

largest, dominant edifice of 2b type andesites (block-and-
ash flow breccias), additional 2c type andesites (subvolca-
nic bodies, lava domes) and subordinate 2d type basaltic
andesites (lava flows), can be well constrained by com-
bined paleomagnetic and K/Ar data. Two new sites at
Rocks of Vadálló-kövek, that is the highest preserved part
of the volcano, have yielded identical paleomagnetism –
reverse polarity with non-rotated declination – and the
same characteristics have been pointed out for the related
Szent Mihály Hill epiclastic breccias and summit lava
dome as well. Paleomagnetism of the basaltic andesite of
Prépost Hill lava flow is also the same. K/Ar ages obtained
on whole rock samples from the Vadálló-kövek area show
15.2—15.4 Ma, well supported by other dates of petrograph-
ically identical rocks in the vicinity (e.g. Visegrád Castle
Hill upper level – 15.3 Ma; Dobogókő Hill – 15.35 Ma;
etc. Table 3). The 2b—c type andesite volcanism of 15 Ma
age was coeval with a widespread effusive lava dome activi-
ty in the S Börzsöny (Karátson et al. 2000). As mentioned
above, in the latter mountains it was a longer lasting erup-
tive stage (16.0 to 14.5 Ma). In contrast, on the basis of our
volcanological and geochronological data, the dominant
2b—c type andesite volcanism of the Visegrád Mts (actually
that of Keserűs Hill) was much shorter. It should have termi-
nated most likely 15 Ma or no later than 14.5 Ma ago, at
end of a reverse polarity zone.

On the other hand there is some uncertainty regarding

the very extinction of the volcanism. There are a few K/Ar
ages younger than 14 Ma, most of them obtained on 2c
andesites and limited to the NE-E margin of the moun-
tains. Another group of young ages on 2d? basaltic andes-
ites comes from the Pilisszentlászló borehole, which seem
to be slightly rejuvenated. However, some other, typical
occurrences of the 2d basaltic andesite (which is otherwise
subordinate in the mountains), namely the Tövises and
Prépost Hill lava flows, are related to the Keserűs Hill vol-
cano and fall into the 3rd paleomagnetic stage. Moreover,
none of the paleomagnetic sites have shown normal polar-
ity without rotation, which is the successive, final paleo-
magnetic stage (14.5—13.7 Ma) in the High Börzsöny
Lava Dome complex (Karátson et al. 2000). Therefore, al-
though some young eruptions, such as a small-scale 2c or
2d type andesitic activity, could have occurred, this
should have been very limited. In contrast, given the
widespread non-rotated normal polarity paleomagnetism
of the High Börzsöny, a voluminous late-stage activity in
the Börzsöny Mts lasted considerably longer.

8.2 Magmatic evolution

In the magmatic evolution of the Visegrád Mts there is

a systematic change from a more (early stage dacites) to a
less silicic (late stage 2d type basaltic andesite) magma,
similarly to the Börzsöny Mountains. During the early

560

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

stage magmatism, mostly dacites and rhyodacites were
formed, usually containing garnet phenocrysts (Harangi et
al. 2001). These rocks show a marked depletion in the Y
and heavy rare earth elements (Fig. 5b) suggesting early
crystallization and fractionation of garnets.

In the main (2nd and 3rd) volcanic phase mostly andes-

ites were formed with variable phenocryst contents. In
general, amphibole is dominant in most of the rock types
accompanied by variable amounts and types of pyroxenes.
In the late stage basaltic andesites, however, amphibole is
lacking, or occurs as subordinate inherited phase. This in-
dicates that during the evolution of the Visegrád volcanic
complex, magmas with decreasing water content were
erupted. Within the amphiboles different groups were
identified in the four different andesite types mostly on
the basis of their oxidization state. In addition, they show
a wide range of composition (from Mg-hastingsite in 2a
type biotite amphibole andesite and 2c type pyroxene am-
phibole andesite to tschermakite in 2b type pyroxene am-
phibole andesite) suggesting crystallization at different
temperatures and pressures. The incompatible trace ele-
ment ratios of the andesites form linear trends suggesting
cogenetic relationships, that is a common source region
and variable degrees of magmatic differentiation. In addi-
tion to the fractional crystallization processes, mixing of
magmas formed at different stages of evolution could also
occur as shown by the petrographic observations. Finally,
Harangi (2001) showed that these rocks also have wide
variation of isotopic composition suggesting that various
degrees of assimilation of crustal material could have been
taken place as well. In summary, the magmatic evolution
of the Visegrád volcanism can be described by erupting of
magma batches deriving from the same source region, but
they underwent variable degrees of magmatic differentia-
tion processes, such as assimilation combined with frac-
tional crystallization and occasionally mixing of magmas.

On the basis of textural properties and mineralogical

composition, we assume a continuous change in magma
generation processes from 2c type andesites to 2d type ba-
saltic andesites. This is also supported by the geochemical
analysis (see Fig. 5).

The generation of late stage basaltic andesite (2d) mag-

ma indicates the end of the volcanic activity. However, we
cannot find a clear magmagenetic succession between the
2a—b—c andesite types. While successive petrographic fea-
tures suggest a 2a—b—c—d magmatic evolution, the 2c
andesite, as seen in the above, may occur in volcaniclastic
deposits of basal stratigraphic position too, and the major-
ity of this andesite type occurs up to the highest strati-
graphic levels (i.e. Ágas Hill post-BFD lava dome). The
exact stratigraphic position of the 2a andesite type is also
ambiguous, although most paleomagnetic and K/Ar data
indicate an early stage (see above).

9. Conclusions

The Middle Miocene volcanic stratigraphy of the Vi-

segrád Mts and its relationship to the neighbouring, co-

eval Börzsöny Mts have been investigated by a com-
bined approach. According to our new stratigraphic mod-
el, the Visegrád volcanism was a three-stage eruptive
activity lasting from 16 to 15.0/14.5 Ma. During the
volcanic period, the initial, shallow submarine eruptions
gave place to an emergent activity in an archipelago. A
comparative drawing about the main volcanic episodes,
indicative K/Ar ages, characteristic fossils, and proposed
informal lithostratigraphic units of the Visegrád Moun-
tains, with respect to main events and geochronology of
the Börzsöny Mountains, is presented in Fig. 6.

Combining field volcanological data with a great

number of new and previous K/Ar ages as well as paleo-
magnetic characters, we propose that the 1st

volcanic

stage was a short-lived (some hundred kyr) dacitic-rhyo-
dacitic activity, started in the S periphery 16 Ma ago.
Mostly in peripheral areas, this activity was preceded by
andesitic ( ± dacitic) phreatomagmatic eruptions too.
Within the dacitic volcanism, less significant, possibly
lava dome-related explosive eruptions occurred earlier,
whereas extrusive and subvolcanic activities may have
lasted somewhat longer ( 16 Ma). The period (ca. 16.5—

16 Ma) proposed for the 1st stage is significantly older

than 15.2—14.8 Ma in a recent model (Korpás (Ed.)
1998), and largely corresponds to the 1st volcanic stage
of the Börzsöny Mts (16.5—16.0 Ma, Karátson et al.
2000). The dacitic pyro- and mostly epiclastic rocks em-
placed during this stage exhibit normal polarity with ca.
30º CCW rotation, whereas the dacitic lava dome activi-
ty should have continued into a reverse polarity zone.

The much more widespread exlusively andesitic volca-

nism of the Visegrád Mts may have followed after a time
gap, at least in the S where interbedded fossil-bearing
sediments indicate a shallow submarine environment.
The renewed activity could have been more or less con-
tinuous, but can be divided into two stages on the basis
of paleomagnetic characters: 2nd stage with reverse po-
larity and ca. 30º CCW declination, and 3rd stage with
reverse polarity and no rotation. The products of these
two stages can be found in the Börzsöny Mts as well,
where the volcanic period was inferred to represent a sig-
nificant rotation event (Karátson et al. 2000). In the Vi-
segrád Mts, the timing of the rotation, during the main
andesitic activity, can be better constrained by the K/Ar
method: 15.5—15.3 ( ± 0.5) Ma. Accordingly, we propose
that the rotation period (that occurred toward the end of
the 16—15 Ma reverse paleomagnetic zone) divides the
2nd and 3rd paleomagnetic stage both in the Börzsöny
and Visegrád Mts.

The rocks of these latter stages have a great variety of

massive (1) and volcaniclastic (2) rocks. (1) Biotite am-
phibole andesites to pyroxene amphibole andesites and
basaltic andesites occur as subvolcanic bodies (includ-
ing sills and dykes), deeply eroded, scattered lava domes,
and lava flows. (2) The petrographically identical volca-
niclastic rocks include block-and-ash flow breccias and
their reworked counterparts (being the most widespread)
as well as deposits of the debris-flow/flood-flow/fluvial
streamflow continuum. All of these rocks were emplaced

561

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

Fig. 6. Geochronology, main volcanic events and proposed lithostratigraphic units of the Visegrád Mts in comparison to the Börzsöny
Mts (Karátson et al. 2000).

subaerially. The main eruptive centre was the Keserűs
Hill amphibole andesite lava dome complex (Karátson et
al. 2006) with a relatively explosive character and minor
lava flows. It underwent a horseshoe-shaped caldera for-
mation toward the N; an outer caldera (Dobogó-Kő Hill)
suggested in the previous literature is not supported.
Around the volcano the volcaniclastic deposits show
systematic changes from proximal to distal facies (e.g.
from pyroclastic breccias to debris-flow deposits etc.).

In the Visegrád Mts there are no indications for a 4th

paleomagnetic stage as is the case with the Börzsöny
Mts. However, with respect to sporadic K/Ar ages, a
younger than 14.5 Ma volcanic activity is likely, but it
should have been very limited in area. In fact, this activi-
ty is constrained in time by the deposition of the Lang-
hian (in terms of Central Paratethys, Lower Badenian)

“Leitha” limestone. In accordance with this, we propose
that the main eruptive period of the Visegrád Mts termi-
nated at 15(—14.5) Ma. This final activity may have been
coeval with the buildup of the High Börzsöny basaltic
andesite lava dome complex.

Acknowledgments: We dedicate this paper to late László
Korpás, who worked in the Börzsöny—Visegrád Moun-
tains for many years. He was always keen to share his ex-
perience with his colleagues, and gave an excellent
review on this work. Ioan Seghedi and Jaroslav Lexa are
also acknowledged for their thorough reviews, and
Miklós Lantos for giving his unpublished paleomagnetic
data on Csódi Hill. D.K. and E.M. thank for the financial
support of Hungarian National Grants OTKA T043664
and T043737. Part of the work was done during D.K.’s

562

KARÁTSON, OLÁH, PÉCSKAY, MÁRTON, HARANGI, DULAI, ZELENKA and KÓSIK

Bolyai and Fulbright scholarships. Some short field trips
visiting the fossiliferous sedimentary and volcanic sedi-
mentary formations were supported by OTKA T49224
and A.D.’s Bolyai scholarship.

References

Báldi T. & Kókay J. 1970: The fauna of the Kismaros tuffite and

the age of the Börzsöny andesitic volcanism. Földt. Közl.
(Bull. Hung. Geol. Soc.) 100, 274—283 (in Hungarian with
German abstract).

Balla Z., Czakó T. & Korpás L. 1977: Report on the preparations of

the geological research in the Dunazug Mountains. Manu-
script, Open Files Hung. Geol. Inst., Budapest (in Hungarian).

Balla Z. & Márton-Szalay E. 1979: Magnetostratigraphy of the

Börzsöny and Dunazug Mountains. Geofizikai Közlemények
(Geophysical Letters) 26, 57—77 (in Hungarian with English
abstract).

Balogh K. 1977—1979: K/Ar ages of rocks of the Börzsöny—

Dunazug Mts. Unpublished summary, ATOMKI (Nuclear Res.
Inst. Hung. Acad. Sci.), Debrecen (in Hungarian).

Balogh K. 1985: K-Ar dating of Neogene volcanic activity in

Hungary: Experimental technique, experiences and methods
of chronological studies. ATOMKI Reports D/1, Debrecen,
277—288.

Benedek K. 1998: Petrological and geochemical studies in the Vi-

segrád Mts. MSc Thesis, Eötvös Univ., Budapest, 1—78 (in
Hungarian).

Bendő Zs. 2002: Volcanological investigations of Visergrád Castle

Hill and vicinity. MSc Thesis, Eötvös Univ., Budapest, 1—70
(in Hungarian).

Bendő Zs., Gméling K., Badics B. & Izing I. 2001: Ignimbrites in

the ravine of Holdvilág-árok. National Competition of Stu-
dents, XXV Conference, Physics/Earth Sciences/Mathematics,
Pécs, Hungary, 1—55 (in Hungarian).

Bendő Zs. & Korpás L. 2005: How much time is needed for lacco-

lith formation? A new approach based on a case study from
Csódi-hegy, Dunabogdány, Hungary. Acta Geol. Hung. 48, 3,
299—316.

Bohn-Havas M. & Korecz-Laky I. 1980: An Eggenburgian fauna

from Csádri Stream at Felsőbogdány (Dunazug Mountains).
Földt. Közl. (Bull. Hung. Geol. Soc.) 110, 276—283 (in Hun-
garian).

Cholnoky J. 1937: The Dunazug Mountains. Földt. Közl. (Bull.

Hung. Geogr. Soc.) 65, 1—27 (in Hungarian with German
abstract).

Dulai A. 1996: Taxonomic composition and paleoecological fea-

tures of the Early Badenian (Middle Miocene) bivalve fauna
of Szob (Börzsöny Mts., Hungary). Ann. Hist.-Natur. Mus.
Natur. Hung. 88, 31—56.

Fisher R. 1953: Dispersion on a sphere. Proc. R. Soc. London, Ser.

A 217, 295—305.

Fodor L., Csontos L., Bada G., Benkovics L. & Györfi I. 1999:

Tertiary paleostress field and structural evolution: a new syn-
thesis. Geol. Soc. London, Spec. Publ. 156, 295—334.

Gill J.B. 1981: Orogenic andesites and plate tectonics. Springer Ver-

lag, Berlin—Heidelberg—New York, 1—390.

Harangi Sz. 1999: Geochemistry and petrogenesis of the volcanic

rocks of Csódi Hill. Topographia Mineralogica Hungariae
VI, 59—85 (in Hungarian).

Harangi Sz., Korpás L. & Weiszburg T. 1999: Miocene calc-alka-

line volcanism of the Visegrád Mts, Northern Pannonian Ba-
sin. Beih. Eur. J. Mineral., Exkursion A, 11, 2, 14—17.

Harangi Sz., Downes H., Kósa L., Szabó Cs., Thirlwall M.F., Mason

P.R.D. & Mattey D. 2001: Almandine garnet in calc-alkaline
volcanic rocks of the Northern Pannonian Basin (Eastern-Cen-
tral Europe): geochemistry, petrogenesis and geodynamic im-
plications. J. Petrology 42, 10, 1813—1843.

Hegedüs Gy. 1953: Contributions to the geology of Visegrád and

its vicinity. Ann. Rep. Hung. Geol. Inst. 1943, 45—49 (in Hun-
garian).

Hibbard M.J. 1995: Petrography to petrogenesis. Prentice Hall, En-

glewood Cliffs, New Jersey, 1—587.

Höfer A. 2003: Vulkanologische und sedimentologische Untersu-

chungen der neogenen Vulkanoklastite nordwestlich von
Pomáz (Visegrád Mts., Ungarn). MSc Thesis, TU Bergakade-
mie, Freiberg, 1—128.

Karátson D. 1995: Ignimbrite formation, resurgent doming and

dome collapse activity in the Miocene Börzsöny Mountains,
North Hungary. Acta Vulcanol. 7, 107—117.

Karátson D., Márton E., Harangi Sz., Józsa S., Balogh K., Pécskay

Z., Kovácsvölgyi S., Szakmány Gy. & Dulai A. 2000: Volca-
nic evolution and stratigraphy of the Miocene Börzsöny
Mountains, Hungary: an integrated study. Geol. Carpathica
51, 1, 325—343.

Karátson D. & Németh K. 2001: Lithofacies associations of an

emerging volcaniclastic apron in a Miocene volcanic complex:
an example from the Börzsöny Mountains, Hungary. Int. J.
Earth Sci. (Geol. Rdsch.) 90, 776—794.

Karátson D., Németh K., Józsa S. & Borbély E. 2001: An ancient

debris avalanche initiated the river loop? The mystery of the
Danube Bend, Hungary. 11

EUG Congress Strasbourg, Ab-

stract Volume, Section EV02.

Karátson D., Németh K., Székely B., Ruszkiczay-Rüdiger Zs. &

Pécskay Z. 2006: Incision of a river curvature due to exhumed
Miocene volcanic landforms: Danube Bend, Hungary. Int. J.
Earth Sci. (Geol. Rdsch.) 95, 5, 929—944.

Koch A. 1877: Geological description of the Danube Trachyt

Group on the right side of the river, with a preface of the
oro- and hydrographic features. Hung. Acad. Sci., Budapest,
1—299 (in Hungarian).

Koch A. 1879: Geology of the Csódi Hill and its vicinity at Bogdány.

Földt. Közl. (Bull. Hung. Geol. Soc.) 9, 1—14 (in Hungarian).

Konečný V. & Lexa J. 1994: Processes and products of shallow

submarine volcanic activity in Southern Slovakia. IAVCEI
General Assembly Abstract Volume, Ankara, Turkey.

Korpás L. (Ed.) 1998: Explanations to the geological map of the

Börzsöny and Visegrád Mountains. Hung. Geol. Inst., Buda-
pest, 1—216 (in Hungarian with English abstract).

Korpás L., Peregi Zs. & Szendrei Zs. 1967: Petrological and geo-

logical investigation of the northern part of the Dunazug
Mountains. Földt. Közl. (Bull. Hung. Geol. Soc.) 97, 211—223
(in Hungarian).

Korpás L. & Csillag-Teplánszky E. 1999: Geological map of the

Börzsöny-Visegrád Mts. and their surroundings, 1: 50,000.
Hung. Geol. Inst., Budapest.

Kósik Sz. 2005: Investigation of volcaniclastic rocks in the

Visegrád Mts. MSc Thesis, Eötvös Univ., Budapest, 1—64.

Láng S. 1955: Physical geography of the Mátra and Börzsöny.

Akadémiai Kiadó, Budapest, 1—512 (in Hungarian with Rus-
sian and German abstract).

Lengyel E. 1953: Geology of the Dunazug andezite mountains.

Ann. Rep. Hung. Geol. Inst. 1951, 17—29 (in Hungarian).

Lexa J. & Konečný V. 1974: The Carpathian volcanic arc: a discus-

sion. Acta Geol. Hung. 18, 3—4, 279—293.

Márton E. & Márton P. 1995: Large-scale rotations in North Hun-

gary during the Neogene as indicated by palaeomagnetic data.
In: Morris A. & Tarling D.H. (Eds.): Palaeomagnetism and
tectonics of the Mediterranean Region. Geol. Soc. London,
Spec. Publ. 105, 153—173.

563

MIOCENE VOLCANISM IN THE VISEGRÁD MOUNTAINS (HUNGARY)

Márton E. & Pécskay Z. 1998: Complex evaluation of paleomag-

netic and K/Ar isotope data of the Miocene ignimbritic volca-
nics in the Bükk Foreland, Hungary. Acta Geol. Hung. 41, 4,
467—476.

Méhes K. 1942: Geological studies in the vicinity of Csódi Hill at

Dunabogdány. Relationes Ann. Inst. Geol. Publ. Hung. B,
Disputationes, 59—93.

Moissette P., Dulai A. & Müller P. 2006: Bryozoan faunas from the

Middle Miocene of Hungary: biodiversity and biogeography.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 233, 300—314.

Moissette P., Dulai A., Escargyel G., Kázmér M., Müller P. &

Saint Martin J.-P. 2007: Mosaic of environments recorded by
bryozoan faunas from the Middle miocene of Hungary.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 252, 530—566.

Müller P. 1984: Decapod Crustacea from the Badenian. Geol.

Hung., Ser. Palaeontologica 42, 1—317.

Pearce J.A. & Parkinson I.J. 1993: Trace element models for mantle

melting: application to volcanic arc petrogenesis. In: Prichard
H.M., Alabaster T., Harris N.B.W. & Neary C.R. (Eds.): Mag-
matic processes and plate tectonics. Geol. Soc. London, Spec.
Publ. 76, 373—403.

Pécskay Z., Lexa J., Szakács A., Balogh K., Seghedi I., Konečný

V., Kovács M., Márton E., Kaličiak M., Széky-Fux V., Póka
T., Gyamati P., Edelstein O., Rosu E. & Žec B. 1995: Space
and time distribution of Neogene-Quaternary volcanism in the
Carpatho-Pannonian region. In: Downes H. & Vaselli O.
(Eds.): Neogene and related magmatism in the Carpatho-Pan-
nonian Region. Acta Vulcanol. 7, 2, 15—28.

Pécskay Z. & Molnár F. 2002: Relationships between volcanism

and hydrothermal activity in the Tokaj Mts., NE Hungary,
based on K-Ar ages. Geol. Carpathica 53, 5, 303—314.

Peters K.F. 1869: Geologische Studien aus Ungarn. II. Die Umge-

bung von Visegrád, Gran, Totis und Zsámbék. Jb. K.-Kön.
Geol. Reichsanst. 10, 4.

Póka T., Zelenka T., Seghedi I., Pécskay Z. & Márton E. 2004: Mi-

ocene volcanism of the Cserhát Mts (N Hungary): integrated
volcano-tectonic, geochronologic and petrochemical study.
Acta Geol. Hung. 47, 2—3, 221—246.

Schafarzik F. & Vendl A. 1929: Geological field trips in the vicini-

ty of Budapest. Stadium Sajtóvállalat Rt., Budapest, 1—343 (in
Hungarian).

Scholz G. 1970: The Tortonian coral fauna of Fekete Hill at Viseg-

rád. Földt. Közl. 100, 192—206 (in Hungarian with English
abstract).

Steiger R.H. & Jäger E. 1977: Subcomission on Geochronology:

Convention on the use of decay constants in geology and geo-
chronology. Earth Planet. Sci. Lett. 36, 3, 359—362.

Szabó Cs., Harangi Sz. & Csontos L. 1992: Review of Neogene and

Quaternary volcanism of the Carpathian-Pannonian Region.
In: Ziegler P.A. (Ed.): Geodynamics of rifting. Case studies on
rifts: Europe and Asia. Tectonophysics 208, 243—256.

Székely B. & Karátson D. 2004: DEM-based volcanic geomorpholo-

gy as a tool for reconstructing volcanic edifices: examples from
the Börzsöny Mts, North Hungary. Geomorphology 63, 25—37.

Vass D. & Balogh K. 1989: The period of Main and Late Alpine Mo-

lasses in the Carpathians. Z. Geol. Wiss. (Berlin) 17, 849—858.

Wein Gy. 1939: Über die geologische Verhältnisse der Umgebung

von Szentendre. Földt. Közl. (Bull. Hung. Geol. Soc.) LXIX,
1—3, 1—27 (in Hungarian with German abstract).

Zelenka T. 1960: Petrological and geological studies in the SW part

of the Dunazug Mountains. Földt. Közl. (Bull. Hung. Geol.
Soc.) 90, 83—102 (in Hungarian).