ArticlePDF Available

Formation of lava stalactites in the master tube of the 1792-1793 flow field, Mt. Etna (Italy)


Abstract and Figures

Lava tubes are often coated with spectacular lava stalactites that are thought to form by a process of lava remelting. Here, we present results from lava stalactites collected inside a master lava tube that fed the 1792–793 Etna flank eruption, which show features rather different from their Hawaiian or Icelandic counterparts. We analyzed three types of stalactites recognized at Mt. Etna on the basis of their morphology, and compared their features with those of the lava flow hosting the tube. Three-dimensional morphologic analyses by SEM, petrographic observations, and mineral and glass composition measured by SEM-EDS, allowed us to infer processes and conditions of stalactite formation. Our results indicate that in all the analyzed stalactites, the nature, abundance and composition of phenocrysts is similar to that of the host lava flow. This finding suggests a common mechanical origin for different types of stalactites, caused by drainage of the tube and dripping of fluid lava from the roof. However, the composition of interstitial glass is significantly different from that of the glassy groundmass measured in historical volcanic rocks of Mt. Etna and suggests that, once stalactites solidified, they were affected by a process of partial melting. Partial melting involved between 12 and 25% of the bulk rock, causing the wide compositional variation and enrichment in K2O measured in our samples.
Content may be subject to copyright.
American Mineralogist, Volume 90, pages 1413–1421, 2005
0003-004X/05/0809–1413$05.00/DOI: 10.2138/am.2005.1760 1413
Stalactites are one of the most spectacular features observed
within lava tubes. Some very colorful and delicate stalactites
form from various mineral precipitates at the end of eruptions
and disappear after some period of time (Giudice and Leotta
1995). However, the majority of stable, long-lasting lava sta-
lactites result from thermal and mechanical processes during
ß ow and drainage of lava through a lava tube. Lava stalactites
on Etna have shapes and sizes that differ signiÞ cantly from their
Hawaiian or Icelandic counterparts. As an example, the delicate,
worm-like structures commonly found on the roofs of lava tubes
in Hawaii (Jaggar 1931; Kauahikaua et al. 1998) have never been
observed on Etna. Features observed within lava tubes in Iceland
suggest dribbling of molten lava from the roof of the tube onto a
rigid crust while the ß ow was still moving slowly (Allred 1998;
Allred and Allred 1998). On Etna, the features of stalactites and
the absence of stalagmites are probably due to the occurrence
of a more viscous lava than in other volcanoes.
On Etna, Calvari and Pinkerton (1999) recognized three
kinds of lava stalactites, distinguishable by their morphology,
of which excellent examples embellish the walls and roof of
the Tre Livelli and Cassone tubes in the 1792–1793 lava ß ow
Þ eld. Stalactites with very smooth surfaces (hereafter: “smooth
stalactites”) form on ridges that are elongated in the ß ow direc-
tion. They are typically red in color, and are believed to form
by re-melting due to gases accumulating below the roof (Jaggar
1931; Kauahikaua et al. 1998; Chadwick 2003). On Etna, smooth
stalactites are typically up to a few centimeters long and at most
2 cm wide at the base, and are conical in shape (Fig. 1). Another
type of stalactite is rough, gray in color, spiky (hereafter: “spiky
stalactites”), and was previously recognized by Jaggar (1931).
On Etna, spiky stalactites are generally less than 0.5 cm wide, a
few centimeters long, and are considered to have formed when
lava completely Þ lled a tube and then drained, either partially or
completely (Calvari and Pinkerton 1999). The resulting stalac-
tites record the dripping of lava from the roof (Fig. 2). The third
type of stalactite, which is not very common, is characterized by
bulbous shapes (hereafter: “bulbous stalactites”). Thin sections
reveal that they are composed of multiple layers of lava with an
external boundary marked by a very thin rind of oxides. Cal-
vari and Pinkerton (1999) interpreted these stalactites as having
been repeatedly coated by lava during ß owage inside the tube.
Our results differ from previous works because we Þ nd that all
types of stalactites share a mechanical origin. Once they cool
and become rigid, stalactites might suffer further complicated
physical and chemical processes (e.g., reheating, remelting, oxi-
dation, lava immersion, etc.) that modify morphology, texture,
and chemical composition. In this paper, we present a study
of the morphology, petrography, mineral and interstitial glass
compositions of different stalactites collected from the master
tube (the Tre Livelli Tube) of the 17921793 lava ß ow Þ eld on
Etna, and compare these data with those measured in the lava
ß ow hosting the tube. This approach allowed us to evaluate the
role of different processes (mechanical, chemical) in the forma-
tion of stalactites.
* E-mail:
Formation of lava stalactites in the master tube of the 1792–1793 ß ow Þ eld, Mt. Etna
1Istituto Nazionale di GeoÞ sica e Vulcanologia – Sezione di Catania (INGV-CT), Piazza Roma 2, 95123 Catania (Italy)
2Istituto Nazionale di GeoÞ sica e Vulcanologia –Sezione Sismologia e TettonoÞ sica, Centro per la Modellistica Fisica e Pericolosità dei Processi
Vulcanica, via della Faggiola 32, 56126 Pisa (Italy)
Lava tubes are often coated with spectacular lava stalactites that are thought to form by a process
of lava remelting. Here, we present results from lava stalactites collected inside a master lava tube
that fed the 17921793 Etna ß ank eruption, which show features rather different from their Hawai-
ian or Icelandic counterparts. We analyzed three types of stalactites recognized at Mt. Etna on the
basis of their morphology, and compared their features with those of the lava ß ow hosting the tube.
Three-dimensional morphologic analyses by SEM, petrographic observations, and mineral and glass
composition measured by SEM-EDS, allowed us to infer processes and conditions of stalactite forma-
tion. Our results indicate that in all the analyzed stalactites, the nature, abundance and composition of
phenocrysts is similar to that of the host lava ß ow. This Þ nding suggests a common mechanical origin
for different types of stalactites, caused by drainage of the tube and dripping of ß uid lava from the
roof. However, the composition of interstitial glass is signiÞ cantly different from that of the glassy
groundmass measured in historical volcanic rocks of Mt. Etna and suggests that, once stalactites so-
lidiÞ ed, they were affected by a process of partial melting. Partial melting involved between 12 and
25% of the bulk rock, causing the wide compositional variation and enrichment in K2O measured in
our samples.
THE 1792–1793 ERUPTION
The 1792–1793 eruption on Mt. Etna is one of the best
examples of emplacement of a compound ß ow Þ eld (Fig. 3)
where several lava tubes at various distances from the source
have been surveyed. The eruption started on 11 May 1792 with
overß ows from the summit crater and continued with lava ß ows
into the Valle del Bove (FF1 in Fig. 3A), a broad valley cutting
the east ß ank of the volcano. The eruptive Þ ssure propagated
south, producing lava ß ows on the southern ß ank of the volcano,
threatening the town of Zafferana about 7 km away (FF2 in Fig.
3A). The eruption ended on 28 May 1793, after one year of
almost continuous activity. The lava ß ow Þ eld extended 5 km
into the Valle del Bove and 6.5 km on its outer, southern ß ank
with a mean thickness of 10 m (Gottini et al. 1980). It covered
a total surface area estimated at 12 km2, and produced a volume
of 120 million cubic meters of lava and 100 000 cubic meters of
pyroclasts (Gottini et al. 1980).
There are seven known lava tubes in this ß ow Þ eld. The Tre
Livelli tube (Fig. 3B) is the highest and longest, and partially
formed within the eruptive Þ ssure. Speleologists discovered an-
other tube (named KTM; Fig. 3B) that is the downhill continua-
tion of the Tre Livelli tube (Corsaro and Giudice 1991; Corsaro et
al. 1995). On the whole, the Tre Livelli–KTM system is among
the longest lava tubes, with difference in elevation between top
and bottom of 450 m and an overall length of about 1750 m. The
stalactites analyzed in this paper were collected from the roof of
the Tre Livelli-KTM system in which most of the original inner
features are still well-preserved.
We focused our study on three samples representative of smooth, spiky, and
bulbous stalactites (Calvari and Pinkerton 1999), collected from the roof of the
Tre Livelli tube. Two samples from the outer surface of the host lava ß ow close
to the skylights (Fig. 3) were also analyzed for comparison. Three-dimensional
morphologic analyses were carried out on stalactites using a Cambridge Stereoscan
360 scanning electron microscope (SEM) from the laboratory of INGV-CT.
We made petrographic observations and measured mineral and glass composi-
tions on samples of both stalactites and host lava with an energy dispersive X-ray
analyzer (LINK eXL) attached to the SEM. Quantitative analyses were processed
using ZAF correction procedures (Armstrong 1988), with natural standards for the
calibration. Analytical conditions were 15 keV accelerating potential and 500 pA
probe current. In order to minimize Na loss, glass compositions were measured
FIGURE 1. Smooth lava stalactites on the roof of the 1792–1793
Tre Livelli tube.
FIGURE 2. Spiky stalactites on the roof of the 1792–1793 Tre Livelli
tube. The bat in the center of roof is about 15 cm high.
FIGURE 3. (A) Sketch map of the complex ß ow Þ eld formed during
1792–1793 eruption on Mt. Etna. FF1 and FF2 = ß ow Þ elds respectively
inside and outside the Valle del Bove. Thick dotted line corresponds to
the rim of the Valle del Bove, a broad valley cutting the east ß ank of
the volcano. A rectangle bounds the area where Tre Livelli-KTM tube
system develops. (B) The development of Tre Livelli-KTM tube system.
Dark gray rectangle corresponds to the area where sampling has been
carried out. (Redrawn after Corsaro et al. (1995); contour line with
elevation in meters.)
using a raster window of about 10 × 10 μm. Replicate analyses of internal natural
standards (mineral and glasses) ensured an analytical precision better than 1% for
SiO2 and Al2O3, about 2% for FeO, MgO, CaO, and K2O, and between 3 and 5%
for the remaining major elements (amount >1%).
Morphologic observations of SEM images show that smooth,
spiky, and bulbous stalactites are locally characterized by dikty-
taxitic texture (Fig. 4). The latter consists of an interconnected
network of microlites separated by irregular angular cavities and
has always been found within the stalactites, at some distance
from the outer rim. The outer surfaces of the stalactites consist of
a thin rind about 50 to 100 μm thick (Fig. 4). Smooth, spiky, and
bulbous stalactites (respectively Figs. 5A, 5B, and 5C) are highly
porphyritic (Porphyritic Index: P.I. = 3236 vol%). Plagioclase is
the most abundant phenocryst (2026 vol%), sub-millimeter to
millimeter in size, euhedral, commonly twinned, and locally with
sieve textures. Clinopyroxene (57 vol%) is euhedral, generally
sub-millimeter in size, and commonly encloses microlites of ox-
ides. Olivine (13 vol%) crystals are generally sub-rounded and
sub-millimeter in size, whereas opaque oxides (1 vol%) are less
common among phenocrysts. The same modal percentages also
have been observed in the two samples of the 1792–1793 host
lava ß ow (see sample 1792-2 in Table 6). The content of vesicles
in bulbous stalactites is higher than in other types (Fig. 5C).
Petrographic observation with the SEM at high magniÞ cation
shows that the thin rind forming the outer surfaces of stalactites
consists of plagioclase, pyroxene, and subordinate glass. Its ex-
terior is marked by aligned oxide microlites (Fig. 6A). The sizes
of crystals forming the rind are signiÞ cantly smaller than those
occurring in the nearby groundmass. The morphology of bulbous
stalactites is due to the overlapping of several layers (Figs. 5C
and 6B) separated by a rind similar to the previously described
one (Fig. 6B). Observations on spiky stalactites show that their
rough surface is generally due to the presence of plagioclase and
pyroxene phenocrysts larger than 2 mm.
Phenocrysts and microlites in smooth, spiky, and bulbous
stalactites have similar compositions. On the whole, the com-
positional range of minerals in stalactites is the same as that
measured in minerals of the 1792–1793 host lava ß ow (Fig. 7).
Selected chemical analyses of minerals are reported in Table 1
(plagioclase), Table 2 (clinopyroxene), Table 3 (olivine), and
Table 4 (oxide). Plagioclase shows normal zoning, with core
compositions ranging from An46 to An86 (average value = An60)
and rims from An35 to An73 (average value = An54). Groundmass
plagioclase spans a compositional range similar to phenocryst
rims (from An39 to An64, average = An52). Clinopyroxene (Fig. 7)
is a diopside-augite (Morimoto 1988). Phenocrysts are weakly
reversely zoned with core compositions ranging from Wo43-48En31-
43Fs11-21 and rims slightly depleted in Fe (Wo 44-48En36-43Fs11-15).
Groundmass microlites (Wo45-48En33-42Fs13-20) overlap the compo-
sitional range of both cores and rims measured in phenocrysts.
Olivine compositions are relatively homogeneous with only
FIGURE 4. Three dimensional view (SEM image) of a smooth
stalactite. The inner part of stalactite shows diktytaxitic texture (see
text). The outer surface consists of a thin smooth rind, 50 to 100
micrometers thick.
FIGURE 5. Thin sections of stalactites cut parallel to the tube roof.
(A) smooth stalactite; (B) spiky stalactite; (C) bulbous stalactite. Dashed
lines mark the boundary of different overlapping layers.
weak normal zoning from core (Fo6578, average = Fo73) to rim
(Fo6475, average = Fo69). Oxides occur as microphenocrysts and
microlites; they are Ti-magnetite, with a variable composition
from Usp24 to Usp29.
The interstitial glass compositions have been analyzed in
smooth, spiky, and bulbous stalactites. Measurements were
carried out along cross sections, in order to investigate com-
positional variations passing from the rind to an intermediate
zone and Þ nally to the inner zone of the stalactite (Table 5). In
bulbous stalactites, we measured glass composition also in rinds
separating various layers. Analyses have been recalculated to
100 wt% to be compared more easily; the unmodiÞ ed totals are
shown in brackets in Table 5.
On the whole, the measurements suggest random variability
(Table 5). In fact, when we plot the compositions of stalactite
glasses in a diagram representing the relative percentages of
oxides (SiO2, Na2O, and K2O; Fig. 8) normally enriched in a
residual liquid, we observe a distribution of data unrelated to the
type (smooth, spiky, or bulbous; Fig. 8a) or to the position (rind,
intermediate, or external zone; Fig. 8b). Due to the fact that the
groundmass of the 1792–1793 lava ß ow is entirely crystallized,
it was impossible to compare the interstitial glass measured in
stalactites with its equivalent in the host lava ß ow. The only
comparison (Table 5) can be made with the interstitial glasses
TABLE 1. Selected chemical analyses of plagioclase measured in 1792-1793 lava fl ow and stalactites
specimen lava ow lava fl ow lava fl ow lava fl ow lava fl ow lava fl ow lava fl ow lava fl ow lava fl ow stalactite stalactite
type smooth smooth
SiO2 52.18 47.95 54.22 52.74 50.72 52.93 51.77 50.87 53.77 49.71 48.63
Al2O3 29.89 32.50 27.97 29.06 30.77 29.55 29.40 26.78 28.05 31.10 32.48
FeOt 0.57 0.54 0.71 0.99 0.83 0.76 1.19 4.85 1.58 1.25 0.51
CaO 13.24 16.15 10.82 11.05 14.32 10.90 12.43 11.15 10.31 14.21 15.77
Na2O 3.60 2.27 5.07 4.72 3.35 4.93 4.68 5.69 5.52 3.19 2.65
K2O 0.37 0.14 0.48 0.67 0.17 0.64 0.19 0.46 0.37 0.25 bdl
Total 99.85 99.56 99.27 99.22 100.15 99.71 99.66 99.80 99.60 99.71 100.04
Cations/O= 8 8 8 8 8 8 8 8 8 8 8
Si 2.37 2.21 2.47 2.41 2.31 2.41 2.37 2.35 2.45 2.28 2.23
Al 1.60 1.77 1.50 1.57 1.65 1.58 1.58 1.46 1.51 1.68 1.75
Fe 0.02 0.02 0.03 0.04 0.03 0.03 0.05 0.19 0.06 0.05 0.02
Ca 0.65 0.80 0.53 0.54 0.70 0.53 0.61 0.55 0.50 0.70 0.77
Na 0.32 0.20 0.45 0.42 0.30 0.43 0.41 0.51 0.49 0.28 0.24
K 0.02 0.01 0.03 0.04 0.01 0.04 0.01 0.03 0.02 0.01 0.00
An 65.54 79.06 52.60 54.19 69.58 52.96 58.84 50.68 49.71 70.06 76.68
Ab 32.26 20.11 44.63 41.90 29.46 43.34 40.09 46.80 48.16 28.46 23.32
Or 2.20 0.83 2.77 3.91 0.96 3.70 1.07 2.51 2.12 1.47 0.00
Notes: Abbreviations: C = core, R = rim, M = microlite, bdl = below detection limit.
TABLE 2. Selected chemical analyses of clinopyroxene measured in 1792–1793 lava fl ow and stalactites
specimen lava ow lava fl ow lava fl ow lava fl ow lava fl ow stalactite stalactite stalactite stalactite stalactite stalactite stalactite
type smooth smooth spiky spiky spiky bulbous bulbous
SiO2 47.41 50.58 47.71 49.88 48.78 49.94 48.01 50.35 49.46 48.85 47.38 48.85
TiO2 1.80 1.34 1.56 1.39 1.26 1.60 1.81 1.16 1.11 1.51 2.65 2.05
Al2O3 5.37 2.26 5.25 3.48 2.67 3.75 5.02 3.69 4.65 2.93 6.45 4.95
FeOt* 8.98 7.96 8.21 7.90 9.29 9.20 9.16 8.44 8.78 10.59 9.28 8.10
MnO bdl bdl bdl bdl 0.35 bdl bdl bdl bdl 0.42 0.35 bdl
MgO 13.78 14.79 13.84 14.15 14.55 12.80 12.05 13.49 13.44 13.52 12.32 12.98
CaO 21.75 22.06 22.25 21.86 21.84 21.42 22.27 21.60 21.45 21.05 20.75 21.40
Na2O 0.92 0.70 0.57 0.88 0.52 0.47 0.55 0.62 0.65 0.60 0.79 1.09
Total 100.01 99.68 99.38 99.54 99.25 99.18 98.87 99.35 99.54 99.47 99.97 99.42
Cations / O= 6 6 6 6 6 6 6 6 6 6 6 6
Site T
Si 1.75 1.87 1.78 1.85 1.82 1.88 1.81 1.88 1.84 1.83 1.77 1.82
Al IV 0.23 0.10 0.22 0.15 0.12 0.12 0.19 0.12 0.16 0.13 0.23 0.18
Fe3+ 0.01 0.03 0.00 0.00 0.06 0.00 0.00 0.00 0.00 0.04 0.00 0.00
Site M1
Al VI 0.00 0.00 0.01 0.00 0.00 0.05 0.04 0.04 0.05 0.00 0.05 0.04
Fe3+ 0.21 0.10 0.17 0.13 0.14 0.02 0.08 0.06 0.10 0.12 0.09 0.11
Ti 0.05 0.04 0.04 0.04 0.04 0.05 0.05 0.03 0.03 0.04 0.07 0.06
Mg 0.74 0.82 0.77 0.78 0.81 0.72 0.68 0.75 0.75 0.76 0.69 0.72
Fe2+ 0.00 0.04 0.01 0.04 0.01 0.17 0.15 0.12 0.08 0.08 0.10 0.08
Site M2
Mg 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Fe2+ 0.05 0.07 0.07 0.07 0.08 0.10 0.06 0.09 0.10 0.10 0.10 0.07
Mn 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.01 0.01 0.00
Ca 0.86 0.88 0.89 0.87 0.87 0.86 0.90 0.86 0.86 0.85 0.83 0.85
Na 0.07 0.05 0.04 0.06 0.04 0.03 0.04 0.04 0.05 0.04 0.06 0.08
Wo 45.71 45.79 46.44 45.81 45.38 46.15 48.22 46.00 45.63 44.25 45.69 46.74
En 40.30 42.73 40.19 41.27 42.07 38.38 36.30 39.98 39.79 39.55 37.75 39.45
Fs 13.99 11.48 13.37 12.92 12.55 15.47 15.48 14.03 14.58 16.19 16.56 13.81
Notes: bdl = below detection limit.
Abbreviations: C = core, R = rim, M = microlite. Cations formula calculated following Morimoto (1988).
of Etnean volcanics erupted from 1329 to 2001 (see compila-
tion in Pompilio et al. 1998; Corsaro and Pompilio 2004; and
Taddeucci et al. 2004). In general, glasses in the stalactites are
signiÞ cantly enriched in SiO2 and Alkalis (especially K2O) and
depleted in TiO2, MgO, CaO, FeO relative to historical Etnean
glasses (Corsaro and Pompilio 2004).
Stalactite glasses: liquids produced by fractional crystal-
Interstitial glasses from the 1983 to 2001 eruptions of Mt.
Etna, show an inverse correlation between CaO/Al2O3 ratio and
K2O. This trend (shaded area in Fig. 9) represents the liquid line
of descent controlled by crystallization of plagioclase, clino-
pyroxene, and olivine at atmospheric pressure (Corsaro and
Pompilio 2004; Taddeucci et al. 2004). In the same diagram,
FIGURE 6. (A) SEM image of a smooth stalactite. The thin rind
consists of pl = plagioclase; cpx = clinopyroxene; ox = oxide. The exterior
of the rind is marked by aligned microlites of oxides. Petrography of the
rind is also the same in spiky and bulbous stalactites. (B) SEM image
of a bulbous stalactite; dashed lines bound the rind separating different
layers. Pl = plagioclase; cpx = clinopyroxene; ox = oxide.
FIGURE 7. Compositions of minerals (plagioclase, clinopyroxene,
olivine, and oxide) observed in smooth, spiky, and bulbous stalactites.
Gray-colored areas deÞ ne the compositional Þ elds of the same minerals
measured in 1792–1793 host lava.
FIGURE 8. Glasses compositions (SiO2, Na2O, and K2O normalized
to 100%) measured in stalactites. Measurements are carried out to
distinguish: (a) smooth, spiky or bulbous stalactites; (b) rind, intermediate
zone or external zone of stalactites cross sections. There are no systematic
variations in glass composition among rind, intermediate, or external
zone nor among smooth, spiky, or bulbous stalactites.
stalactite stalactite stalactite stalactite stalactite stalactite
smooth smooth spiky spiky spiky bulbous
54.44 53.85 49.11 52.48 53.58 52.01
27.61 28.21 31.69 29.39 28.17 29.45
0.83 1.23 0.71 1.11 1.32 1.01
11.38 10.68 14.63 12.35 11.21 12.73
4.77 5.43 3.28 4.52 5.40 4.62
0.53 0.31 0.20 0.21 0.21 0.11
99.56 99.71 99.62 100.06 99.89 99.93
8 8 8 8 8 8
2.48 2.45 2.26 2.38 2.44 2.37
1.48 1.51 1.72 1.57 1.51 1.58
0.03 0.05 0.03 0.04 0.05 0.04
0.55 0.52 0.72 0.60 0.55 0.62
0.42 0.48 0.29 0.40 0.48 0.41
0.03 0.02 0.01 0.01 0.01 0.01
55.13 51.16 70.32 59.43 52.80 59.99
41.82 47.07 28.53 39.36 46.02 39.40
3.06 1.77 1.14 1.20 1.18 0.62
TABLE 1. —Extended
the average value of stalactite glasses plots in an area that could
be the extension of the liquid line.
To verify this hypothesis, we calculated the liquid line-of-
descent for the 1792–1793 lava at atmospheric pressure using the
program MELTS (Ghiorso and Sack 1995). For this purpose, we
used the groundmass of the 1792−1793 lava ß ow as the starting
composition, calculated for sample 1792-2 by subtracting the
abundance of phenocrysts from the whole rock (Table 6). The
need to calculate this parameter was due to the lack of glass in
the groundmass of the 17921793 lava ß ow, which was com-
pletely crystallized. Calculations were carried out for a range of
temperatures between 850 and 1110 °C and for Ni-NiO buffer
conditions. The liquid line-of-descent modeled for decreasing
temperatures at atmospheric pressure shows the highest K2O
value at about 1000 °C. Below this temperature, the abundance
of K2O in the melt decreases abruptly. Consequently, the modeled
TABLE 3. Selected chemical analyses of olivine measured in 1792–1793 lava fl ow and stalactites
specimen lava ow lava fl ow lava fl ow lava fl ow stalactite stalactite stalactite stalactite stalactite stalactite
type smooth smooth spiky spiky bulbous bulbous
SiO2 38.07 37.92 38.50 36.03 38.75 37.38 37.95 38.45 38.31 36.92
FeOt 23.33 24.37 20.02 31.71 20.22 24.52 24.32 25.03 24.95 29.80
MnO 0.42 0.52 0.30 0.90 0.51 0.67 0.46 0.42 0.65 0.98
MgO 37.93 37.19 40.58 30.99 40.56 37.24 36.85 35.43 35.98 32.05
CaO 0.36 0.32 0.23 0.57 0.20 0.17 0.35 0.37 0.15 0.41
Total 100.10 100.31 99.63 100.20 100.24 99.98 99.93 99.70 100.04 100.16
Cations/O= 4 4 4 4 4 4 4 4 4 4
Si 1.00 0.99 1.00 0.99 1.00 0.99 1.00 1.02 1.01 0.99
Fe 0.51 0.53 0.43 0.73 0.43 0.54 0.54 0.55 0.55 0.71
Mn 0.01 0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.01 0.02
Mg 1.48 1.45 1.56 1.26 1.56 1.47 1.45 1.40 1.41 1.27
Ca 0.01 0.01 0.01 0.02 0.01 0.00 0.01 0.01 0.00 0.01
Fo 74.35 73.12 78.33 63.53 78.15 73.03 72.98 71.62 71.98 64.13
Notes: Abbreviations: C = core, R = rim.
TABLE 4. Selected chemical analyses of oxide measured in 1792–1793 lava fl ow and stalactites
specimen lava ow lava fl ow lava ow stalactite stalactite stalactite stalactite stalactite stalactite stalactite
type smooth smooth spiky spiky bulbous bulbous bulbous
SiO2 0.31 bdl 0.37 0.37 bdl 0.31 0.38 0.38 0.28 bdl
TiO2 11.88 12.02 11.79 12.21 12.34 11.70 12.07 13.50 13.26 12.82
Al2O3 6.55 6.03 6.37 5.38 6.10 6.01 5.98 5.10 5.23 6.20
FeOt* 73.60 73.43 74.12 74.74 74.92 74.17 73.89 74.17 73.57 73.85
MnO bdl 0.51 0.52 0.79 0.63 0.38 0.25 0.54 0.71 0.45
MgO 4.98 5.08 4.86 4.84 4.43 5.01 4.95 4.91 5.60 5.15
CaO bdl bdl bdl 0.25 0.23 0.28 0.35 0.15 bdl bdl
Cr2O3 0.26 bdl bdl bdl bdl 0.35 0.44 bdl bdl bdl
Total 97.58 97.06 98.03 98.58 98.65 98.21 98.31 98.75 98.65 98.47
Cations / O= 4 4 4 4 4 4 4 4 4 4
Si 0.01 bdl 0.01 0.01 bdl 0.01 0.01 0.01 0.01 bdl
Ti 0.31 0.32 0.31 0.32 0.32 0.30 0.31 0.35 0.34 0.33
Al 0.27 0.25 0.26 0.22 0.25 0.25 0.24 0.21 0.21 0.25
Fe2+ 2.14 2.15 2.15 2.17 2.17 2.15 2.14 2.15 2.12 2.14
Mn 0.00 0.02 0.02 0.02 0.02 0.01 0.01 0.02 0.02 0.01
Mg 0.26 0.27 0.25 0.25 0.23 0.26 0.26 0.25 0.29 0.27
Ca bdl bdl bdl 0.01 0.01 0.01 0.01 0.01 bdl bdl
Cr 0.01 bdl bdl bdl bdl 0.01 0.01 bdl bdl bdl
Usp % 36.79 35.31 35.64 34.92 36.48 34.61 36.11 38.80 37.09 37.58
Notes: bdl = below detection limit. Ulvospinel content calculated following Stormer (1983).
TABLE 5. Selected chemical analyses of interstitial glasses measured in stalactites compared with average 1983–2001 Etnean interstitial glasses
Stalactite smooth smooth smooth smooth smooth spiky spiky spiky spiky bulbous bulbous bulbous bulbous
SiO2 65.02 59.66 50.05 51.02 61.60 65.76 56.13 62.57 62.87 62.52 60.18 54.19 63.32
TiO2 0.47 bdl 1.63 1.37 bdl bdl 1.81 2.25 1.96 1.17 1.79 1.23 bdl
Al2O3 17.96 23.24 22.26 20.15 18.79 18.88 19.51 17.33 16.96 20.49 18.66 16.70 18.30
FeOt 1.13 1.56 11.21 9.34 3.31 1.02 4.82 1.72 2.89 2.39 3.34 5.82 1.42
MgO bdl bdl 1.32 2.70 bdl bdl 0.47 0.30 bdl bdl 0.36 2.42 bdl
CaO bdl 6.04 1.19 7.15 bdl 0.25 2.23 0.26 bdl bdl bdl 8.08 bdl
Na2O 2.82 5.78 2.67 4.16 1.72 6.05 3.80 2.23 1.93 4.73 2.71 2.63 1.84
K2O 12.59 3.17 8.75 4.10 14.58 8.06 10.31 11.67 11.89 7.14 11.87 7.34 15.13
P2O5 bdl bdl 0.74 bdl bdl bdl 0.45 0.74 0.82 0.84 0.44 1.60 bdl
Cl bdl 0.12 0.18 bdl bdl bdl 0.47 0.92 0.68 0.72 0.66 bdl bdl
S bdl bdl bdl bdl bdl bdl bdl bdl bdl bdl bdl bdl bdl
Total 100 100 100 100 100 100 100 100 100 100 100 100 100
(99.44) (98.53) (98.73) (97.72) (98.12) (98.84) (99.02) (98.64) (99.23) (98.52) (98.15) (98.84) (97.38)
Notes: no. 200 = number of analyses , INN = inner zone of stalactite cross section, INTD = intermediate zone of stalactite cross section, bdl = below detection limit
Analyses have been recalculated to 100, the unmodifi ed totals are bracketed.
path does not overlap the compositional Þ eld of glasses measured
in the stalactites. This result implies that a process other than
fractional crystallization is needed to explain the formation of
interstitial glasses in the stalactites.
Stalactite glasses: products of melting process
An alternative mechanism for the formation of stalactite
glasses is local partial melting of solidiÞ ed stalactite already
existing at the roof of the tube. In this case, the diktytaxitic voids
within the lava stalactites could result from mobilization of the
melted groundmass.
Partial melting of a rock in terms of major-element composi-
tion can be modeled by mixing calculations. In fact, if a linear
mixing calculation is dealing with partial melting, then the
compositions of the source rock (Z), the partial melt removed
from the rock (X), and the residuum (Y) must be related linearly,
following the lever rule suggested by Ragland (1989):
Z = NX + (1 N) Y (1)
where N = fraction melted.
To test our hypothesis, we insert the following compositions
into Equation 1: X = average interstitial glass composition of sta-
lactites (Table 5); Z = 17921793 bulk-rock composition (Table
6). We then solve Equation 1 for Y, constraining N between 0.05
and 0.25. Calculations have been carried out for all oxides, but
we plotted only the results for K2O (Fig. 10). In fact, this oxide
shows a wide variability in the glassy portion of stalactites, and
it represents a useful test for our hypothesis because of its afÞ nity
for the liquid phase rather than for the mineralogical assemblage
observed in 1792–1793 lava.
The approach suggested by Ragland (1989) proves quite
useful because we are forced to stop iterations when negative
values occur for Y, as the restite compositions cannot be nega-
tive. In this way, although we lack measurements of the restite
in lava ß ow, we can constrain the maximum amount of melting
required to explain the K2O content measured in our stalactite
samples. The average composition of interstitial glasses (K2O =
9.74 wt%) measured in stalactites results from a partial melting
process involving about 17% of the bulk rock (continuous line in
Fig. 10). The variability of K2O content measured in the stalactite
glasses (σ = 3.4 7, see Table 5), can be explained by considering
different partial melting conditions. A 12% partial melt of the
bulk rock explains the value of K2O = 13.20 wt% measured in
stalactite glasses (dashed line in Fig. 10), whereas a 25% partial
melting of the bulk rock explains the value of K2O = 6.27 wt%
measured in stalactite glasses (dotted line in Fig.10).
Field evidences and direct observations (Calvari and Pinker-
ton 1999)—compared with textural, mineralogical, and compo-
sitional data presented in this paper—suggest that the smooth,
spiky, and bulbous stalactites that formed during the 17921793
FIGURE 10. Results of partial melting calculation for the 17921793
magma, following the approach suggested by Ragland (1989). Because
a negative K2O content in the restite is not possible, the average
composition of interstitial glasses in stalactites results from a partial-
melting process involving about 17% of the bulk rock (continuous line);
the variability of K2O content really measured in stalactite glasses can be
explained by widening the partial-melting conditions from 12% (dashed
line) to 25% (dotted line). See text for discussion.
FIGURE 9. CaO/Al2O3 vs. K2O for the liquid line-of-descent of the
17921793 magma calculated at atmospheric pressure using the program
MELTS (Ghiorso and Sack 1995). The percentage of residual liquids
remaining after each temperature decrement (from 1110 to 850 °C) is
shown in parentheses. The average composition of stalactite glasses (error
bars = 1σ) and the compositional Þ eld of 1983–2001 Etnean interstitial
glasses are plotted for comparison.
bulbous Average stalactite Average 1983–2001
glasses Etnean glasses
RIND no. 31 σ no. 200 σ
64.48 60.10 4.61 50.84 2.41
0.64 0.80 0.77 2.13 0.27
17.26 18.90 2.20 16.03 0.7
3.69 3.81 2.94 10.29 1.25
bdl 0.53 0.91 2.87 0.88
bdl 1.76 2.73 6.78 1.47
6.19 3.53 1.85 4.61 0.82
7.76 9.74 3.47 3.98 1.03
bdl 0.49 0.67 1.02 0.18
bdl 0.18 0.30 0.18 0.18
bdl 0.02 0.09
Etna eruption all have a mechanical origin, caused by drainage
of the tube and dripping of ß uid lava from the roof. Smooth and
spiky stalactites then derive from a rather simple process that
includes a temporary Þ lling of the tube by lava followed by drain-
ing. Bulbous stalactites result from more complex processes that
involve either repeated cycles of Þ lling and draining of lava, or
splashing of lava ß owing within the tube. These two processes
are responsible for the coating of layers that distinguish the bul-
bous type of stalactites. The fact that bulbous stalactites show a
high content of vesicles and only have been found close to the
eruptive Þ ssure and/or at constrictions of the lava tube, suggests
that they form where lava is still ß uid and contains enough gas
to be able to produce bubbling and splashing. These sites also
favor turbulence and rapid oscillations of lava level that promote
lava splashing.
The surface of a stalactite cools more rapidly than its inner
portion, and this is why every type of stalactite is generally coated
by a rind consisting of plagioclase and pyroxene microlites
smaller than those occurring in the inner groundmass.
Once the mechanical processes described above have formed
a lava stalactite, other processes can occur and be superimposed
on the stalactites. Remelting of the tube roof has been invoked
for the formation of lava straws in Hawaii (Jaggar 1931), but this
process can be ruled out for our Etna samples. In fact, petrogra-
phy of lava stalactites from the Tre Livelli tube system shows that
in all types (smooth, spiky, and bulbous) of stalactites, the nature,
abundance, and composition of phenocrysts of the host lava ß ow
is preserved. This feature indicates that the melting process, if it
occurred, did not advance enough to produce signiÞ cant changes
in the mineral assemblage. On the other hand, when we compare
the compositions of interstitial glasses within stalactites with the
glasses measured in historic volcanic rocks of Etna (Pompilio et
al. 1998; Corsaro and Pompilio 2004; Taddeucci et al., 2004) and
with products of the liquid line-of-descent calculated by numeri-
cal modeling (Fig. 9), it is evident that partial melting occurred
involving between 12 and 25% of the groundmass. This process
appears to be local and random, and affects all the stalactite
types to a similar extent. In addition, a limited amount of partial
melting explains the wide compositional variability of stalactite
interstitial glasses and their enrichment in K2O. In fact, for the
17921793 Etna lava, K is an element with a strong afÞ nity for
the melt phase. The irregular angular cavities between miner-
als forming the groundmass inside the stalactites (diktytaxitic
texture) could derive by re-mobilization of the locally melted
groundmass. The partial melting of already solidiÞ ed stalactites is
caused by a temporary increase in temperature due to lava-level
changes within the tube, and is related to local heat loss. In turn,
the balance between heat produced and lost is related to: (1) the
initial temperature of the lava; (2) the surface crust speed; (3)
the surface slope, which controls crust breakage and then surface
heat loss; (4) the amount of crystallization and then of latent heat
released; and (5) the number and size of skylights (Keszthelyi
1995) along the tube, which cause entrapment of air at ambient
temperature and then temperature decrease inside the tube.
Finally, external portions of a stalactite can remain exposed
for a long time to hot gases that circulate in the tube and which are
progressively mixed with air from the atmosphere. The resulting
gases will be more oxidized than those (QFM-NNO buffer) that
occurr in the magma during intratelluric crystallization. These
oxidizing environmental conditions are responsible of Þ lms of
thin oxides that coat our stalactites samples.
We have analyzed three samples of Etnaʼs lava stalactites
collected within a lava tube of the 1792–1793 eruption. All show
similar features and are compatible with a common mechanism
of formation during drainage of the lava tube. Fluid lava dripping
from the tube roof forms smooth and spiky stalactites. Bulbous
stalactites show, in addition, multiple lava coatings due to splash-
ing of lava within the tube. Interstitial glasses of lava stalactites
are locally enriched in potassium with respect to Etnean historical
glasses. This can be explained by a partial melting between 12
and 25% of the bulk rock, superimposed and subsequent to the
stalactite formation.
The authors are indebted to C.R. Thornber for fruitful scientiÞ c discussion and
for the careful review of an earlier version of this paper. We thank all speleologists
from the Centro Speleologico Etneo in Catania, and especially Marco Liuzzo and
Nino Marino, for their expert Þ eld assistance during our visits to Etnaʼs lava tubes
and collection of the samples here analyzed. P. Asimov, M. Coombs, P. Mazzolei,
and A. Soule are also acknowledged for their comments and suggestions that
helped to improve the paper.
Allred, K. (1998) Lava tube remelt by radiant heat and burning gases. International
Journal of Earth Sciences, 27, 125134.
Allred, K. and Allred, C. (1998) The origin of tubular lava stalactites and other
related forms. International Journal of Speleology, 27, 135145.
Armstrong, J.T. (1988) Quantitative Analysis of Silicate and Oxide Materials:
Comparison of Monte Carlo, ZAF, and ϕ(ρz) Procedures, In D.E. Newbury, Ed.,
Microbeam Analysis-1988, p. 239246. San Francisco Press, California.
Buemi, A. and Pompilio, M. (1992) I prodotti dellʼattività storica dellʼEtna: dati
preliminari su alcune eruzioni del periodo 15361792. Istituto Internazionale
di Vulcanologia. CNR, Open File Report.
Calvari, S. and Pinkerton, H. (1999) Lava tube morphology on Etna and evidence for
lava ß ow emplacement mechanisms. Journal of Volcanology and Geothermal
Research, 90, 263280.
Chadwick, W.W.J. (2003) Quantitative constraints on the growth of submarine lava
pillars from a monitoring instrument that was caught in a lava ß ow. Journal of
Geophysical Research, 108(B11), 2534; doi: 10.1029/2003JB002422.
Corsaro, R.A. and Giudice, G. (1991) Quasi record ai “Tre livelli”. Speleologia,
24, 5659.
Corsaro, R.A. and Pompilio, M. (2004) Magma dynamics in the shallow plumbing
system of Mt. Etna as recorded by compositional variations in volcanics of
TABLE 6. Compositions used in mass balance calculations
sample 1792-2 plagioclase clinopyroxene olivine oxide groundmass
bulk rock * * * * calculated
SiO2 49.02 51.66 50.81 38.06 0.33 48.52
TiO2 1.65 bdl 1.18 bdl 11.98 2.39
Al2O3 18.55 29.95 3.76 bdl 5.70 16.63
Fe2O3 2.92 bdl bdl bdl bdl 4.57
FeO 6.94 0.89 8.37 24.42 76.69 8.01
MnO 0.18 bdl bdl 0.46 0.40 0.26
MgO 4.94 bdl 13.74 36.77 4.90 4.40
CaO 9.75 13.38 21.63 0.28 bdl 7.43
Na2O 3.55 3.79 0.50 bdl bdl 3.99
K2O 1.59 0.33 bdl bdl bdl 2.36
P2O5 0.51 bdl bdl bdl bdl 0.80
1792-2 mineral abundance
pl 25.47
cpx 7.33
ol 3.00
ox 0.33
Notes: 1792-2 composition and mineral abundance is from Buemi and Pompilio
* Mineral composition used for mass balance calculations, bdl: below detec-
tion limit.
recent summit activity (1995–1999). Journal of Volcanology and Geothermal
Research, 137, 5571.
Corsaro, R.A., Giudice, G., and Puglisi, G. (1995) Il sistema 3 livelli-ktm: studio
comparato di una colata con gallerie di scorrimento lavico. I° Convegno
Regionale di Speleologia della Sicilia, I, p. 6676, Ragusa.
Ghiorso, M.S. and Sack, R.O. (1995) Chemical mass transfer in magmatic pro-
cesses IV. A revised and internally consistent thermodynamic model for the
interpolation and extrapolation of liquid-solid equilibria in magmatic sys-
tems at elevated temperatures and pressures. Contribution to Mineralogy and
Petrology, 119, 197212.
Giudice, G. and Leotta, A. (1995) Le alchimie di un vulcano: la Grotta Cutrona.
Speleologia, 33, 1420.
Gottini, V., Rizzo, S., and Sturiale, C. (1980) I fenomeni eruttivi dellʼEtna nel
secoloXVIII. Atti Accademia Gioenia Scienze Naturali, 7, 6770.
Jaggar, T.A. (1931) Lava stalactites, stalagmites, toes, and squeeze-ups. The
Volcano Letter, 345, 13.
Kauahikaua, J., Cashman, K.V., Mattox, T.N., Heliker, C.C., Hon, K.A., Mangan,
M.T., and Thornber, C.R. (1998) Observations on basaltic lava streams in tubes
from Kilauea Volcano, island of Hawaiʼi. Journal of Geophysical Research,
103, 2730327323.
Keszthelyi, L. (1995) A preliminary thermal budget for lava tubes on the Earth and
planets. Journal of Geophysical Research, 100, 2041120420.
Pompilio, M., Trigila, R., and Zanon, V. (1998) Melting experiments on etnean
lavas: the calibration of an empirical geothermometer to estimate the eruptive
temperature. Acta Vulcanologica, 10, 19.
Ragland, P.C. (1989) Basic Analytical Petrology, 369 p. Oxford Univ. Press,
New York.
Taddeucci, J., Pompilio, M., and Scarlato, P. (2004) Conduit processes during the
July-August 2001 explosive activity of Mt. Etna (Italy): inferences from glass
chemistry and crystal size distribution of ash particles. Journal of Volcanology
and Geothermal Research, 137, p. 33–54.
... Differences in these characteristics allowed us to elaborate a model for the plausible and that the predominance of one of the two could be the origin of different types of stalactites. Mechanical origin, caused by drainage of the tube and dripping of lava stuck to the ceiling was invoked by Corsaro et al. (2005) for smooth and spiky stalactites found in lava tubes of the 1792-1793 eruption of Mount Etna; whereas the re-heating and melting process has been considered by other authors (Jaggar, 1931;Kauahikaua and Cashman, 1998), taking into account the reactions developing between magmatic gases and air. With this respect, it is worth considering that the only source of heat able to increase temperature is given by the exothermic reactions of gas combustion as for example between magmatic hydrogen and atmospheric oxygen. ...
... The cave Grotta del Santo (GS) is a lava tube measuring 900 m originated during a prehistoric eruption and consisting of at least 12 minor galleries, diverging from a main central tube. Grotta Cassone (GC) was developed in the 1792-1793 eruption lava field and has been previously described by Calvari and Pinkerton (1999) and Corsaro et al. (2005). This lava tube extends for 300 m and in section shows both simple and compound shapes. ...
Full-text available
Basaltic lava flows can form tubes in response to the cooling of the outer surface. We collected lava stalactites (frozen lava tears) and sampled lava from the ceilings of three lava tubes on Mount Etna. Comparison of the petrographic characters between ceiling lavas and relative stalactites reveal surprising differences in the groundmass textures and crystal compositions. Major and trace element contents in stalactites show only a slight increase in alkali and SiO2 compared to ceiling lava, whereas significant differences exist in composition and textures between plagioclases within the ceiling lava and those within the stalactites, being in the last case definitively more An-rich. We advance the hypothesis that the high temperature reached in the cave caused the exsolution of the volatiles still trapped in the dripping melt. The volatiles, mainly H2O, formed bubbles and escaped from the melt; such a water-loss might have promoted the silicate polymerization in the stalactites resulting in the growth of An-rich plagioclase phenocrysts. Our results have important implications: in fact plagioclase phenocrysts are usually associated with intratelluric growth and are often considered as the main petrologic evidence for the existence of a magma chamber. The textural and chemical features of plagioclases in stalactites prove that phenocrysts growth in syn to post-eruptive conditions is plausible and clearly explain the relatively low viscosity of many phenocrysts-rich lava flows on Mount Etna, as well as on many other volcanoes around the world. Therefore, we can conclude that plagioclase phenocrysts cannot exclusively be considered as having originated within a magma chamber.
... It is not common in the proximal deposits. Compared with lava stalactites (Corsaro et al., 2005), the structure has a rough surface, does not originate from the lava flow, and is probably related to volcanic ash dribbling in the cavity. ...
Full-text available
Large-scale pyroclastic density currents (PDCs) produce inhomogeneity of distributions, thicknesses, stratigraphic structures, and temperatures, thereby causing poor understanding of PDC emplacement processes. Millennium Eruption (ME) of Tianchi volcano in 946 A.D. produced masses of PDCs, covering an area of more than 1570 km2 around the caldera in the Chinese territory. To understand PDC emplacement processes, we examine the proximal to distal variations in the PDC facies during the ME by analyzing the topography, stratigraphy, granulometry, and petrography. The topography of Tianchi volcano has different slopes, including the cone (> 6°), lava shield (2–5°) and lava plateau (< 1°), which affected the PDC emplacement during the ME. The shape of the PDC deposits radiates outward from the caldera, and the periphery has two pyroclastic aprons. The proximal strata (< 10 km) are less than 4 m thick and consist of eutaxitic and lava-like structures in the gullies of the cone. The medial strata (10–20 km) are 10–60 m thick, with columnar joints, pumice-rich layers, and lithic-rich layers in the valleys of the shield. The distal strata (> 20 km) are less than 10 m thick, with coarse-tail layers, ground surge layers, climbing layers, and carbonized woods in the plateau. From the proximal to the distal strata, the emplacement temperature of the PDCs decreased from 740 to 280°C. The medial layers indicate a gravitational differentiation effect and weakened transportation, which led to the formation of thick beds. The distal PDCs demonstrate fluidization characteristics and a density lower than that of the proximal and medial strata, which exhibit clear stratification. Grain size data (< 64 mm) show that the median diameters decreases with increasing distance from the caldera. Rhyolitic and trachytic pumice were found and are assumed to be associated with the composition of the magma chamber. Magmatic mixing structures were confirmed through microscope observations. The volume of the PDCs in the Chinese territory is approximately 7 km3. Our results assist in unraveling some PDCs emplacement processes, marked by a high level of topography confinement, about 50 km runout distance, 280–750°C emplacement temperatures, and more than 7 km3 volume, which provide insights of the PDCs destruction during the ME.
... The InGaAs photodiode with a response at 1300 nm appears to be most sensitive, although the 940 nm Silicon photodiode also shows appreciable fine structure, on time scales that are consistent with appearances of fire fountains. to explain the formation of lava stalactites, i.e., dripping of fluid lava from the roof as the tube drains [38] or melting of the tube roof material from temperatures above the solidus [39]. Our data suggest that the stalactites form in close proximity to a hot, active flow, and are at a temperature of $1000°C, cooler than the lava flowing in the channel and below the basalt solidus temperature. ...
Full-text available
Remote thermal monitoring of active volcanoes has many important applications for terrestrial and planetary volcanic systems. In this study, we describe observations of active eruptions on Kilauea and Erta Ale volcanoes using a short-wavelength, high-resolution, consumer digital camcorder and other non-imaging thermal detectors. These systems revealed brightness temperatures close to the eruption temperatures and temperature distributions, morphologies and thermal structures of flow features, tube systems and lava fountains. Lava flows observed by the camcorder through a skylight on Kilauea had a peak in maximum brightness temperatures at 1230 °C and showed brightness temperature distributions consistent with most rapid flow at the center. Surface brightness temperatures of cooling lava flows on Kilauea were close to 850 °C. Centimeter-scale thermal features are evident around pahoehoe ropes and inflated flows and stalactites in lava tubes. Observations of the fountaining Erta Ale lava lake in February 2011 extend the baseline of observations of the eruptive episode begun in late 2010. We observed a fountain using the camcorder and found a peak in maximum brightness temperatures at 1164 °C, consistent with previous studies. Steep temperature gradients were observed across centimeter-scale distances between the highly exposed fountain and cracks and the much cooler lava lake surface and crater walls. The instrument and methods described here lead to robust pictures of the temperatures and temperature distributions at these volcanoes and reveal desired characteristics of planetary remote sensing platforms for the study of volcanically active bodies such as Io.
... Stalactite formation is attributed to thermal and mech- anical processes during flow and drainage of lava through a lava tube ( Corsaro et al., 2005). These stalactites observed on the roofs of tubes and ephemeral vents, record the dripping of lava from the roof. ...
Full-text available
This study focuses on the compound pahoehoe lava flow fields of the 2000 eruption on Mount Cameroon volcano, West Africa and it comprehensively documents their morphology. The 2000 eruption of Mount Cameroon took place at three different sites (sites 1, 2 and 3), on the southwest flank and near the summit that built three different lava flow fields. These lava flow fields were formed during a long-duration (28th May–mid September) summit and flank eruption involving predominantly pahoehoe flows (sites 2 and 3) and aa flows (site 1). Field observations of flows from a total of four cross-sections made at the proximal end, midway and at the flow front, have been supplemented with data from satellite imagery (SRTM DEM, Landsat TM and ETM+) and are used to offer some clues into their emplacement. Detailed mapping of these lava flows revealed that site 1 flows were typically channel-fed simple aa flows that evolved as a single flow unit, while sites 2 and 3 lava flow fields were fed by master tubes within fissures producing principally tube-fed compound pahoehoe flows. Sites 2 and 3 flows issued from ∼ 33 ephemeral vents along four NE–SW-trending faults/fissures. Pahoehoe morphologies at sites 2 and 3 include smooth, folded and channelled lobes emplaced via a continuum of different mechanisms with the principal mechanism being inflation. The dominant structural features observed on these flow fields included: fissures/faults, vents, levees, channels, tubes and pressure ridges. Other structural features present were pahoehoe toes/lobes, breakouts and squeeze-ups. Slabby pahoehoe resulting from slab-crusted lava was the transitionary lava type from pahoehoe to aa observed at all the sites. Transition zones correspond to slopes of > 10°. Variations in flow morphology and textures across profiles and downstream were repetitive, suggesting a cyclical nature for the responsible processes. Copyright © 2010 John Wiley & Sons, Ltd.
... Physical evidence for the existence of a vapor phase within the voids is preserved on the underside of the upper crust. Lava drips, similar to lava stalactites found in terrestrial lava tubes[13][14][15], occur in most samples collected near collapse features. Lava drips, 1–10 cm in length, are commonly terminated by a bulbous, rounded head (Fig. 1b,c)[5,7]. ...
Evidence for the interaction between seawater and lava during emplacement on the deep seafloor can be observed in solidified flows at a variety of scales including rapid quenching of their outer crusts and the formation of lava pillars through the body of the flow. Recently, an additional interaction, incorporation of heated seawater (vapor) into the body of a flow, has been proposed. Large voids and vesicles beneath the surface crusts of mid-ocean ridge crest lobate and sheet lava flows and lava drips found within those cavities have been cited as evidence for this interaction. The voids resulting from this interaction contribute to the high porosity of the shallow ocean crust and play an important role in crustal permeability and hydrothermal circulation at mid-ocean ridges, and thus it is important to understand their origin. We analyze lava samples from the fast-spreading East Pacific Rise and intermediate-spreading Galapagos Spreading Center to characterize this process, identify the source of the vapor, and investigate the implications this would have on submarine lava flow dynamics. We find that lava samples that have interacted with a vapor have a zone of increased vesicularity on the underside of the lava crust and a coating of precipitate minerals (i.e., crystal fringe) that are distinct in form and composition from those crystallized from the melt. We use thermochemical modeling to simulate the reaction between the lava and a vapor and find that only with seawater can we reproduce the phase assemblage we observe within the crystal fringes present in the samples. Model results suggest that large-scale contamination of the lava by mass exchange with the vapor is unlikely, but we observe local enrichment of the lava in Cl resulting from the incorporation of a brine phase separated from the seawater. We suggest that high eruption rates are necessary for seawater incorporation to occur, but the mechanism by which seawater enters the flow has yet to be resolved. A persistent vapor phase may be important in inhibiting the collapse of lava flow roofs during natural waxing and waning of lava levels during emplacement allowing lava pathways to be maintained during long lived eruptions. In addition, we illustrate the potential for a persistent vapor layer to increase local flow rates within submarine flows by up to a factor of three, thereby influencing how lava is distributed across the ridge crest.
Volcanic centers are complex, dynamic landforms. The stunning morphological variety of volcanic landforms is due to a combination of tectonic setting, eruption style, magma composition and volume, surface environment, and age, i.e., the time the landform has existed and evolved. The considerable variety of volcanic landforms reflects the large variability of these parameters. In turn, volcanoes affect their surrounding landscapes. The influence that volcanism exerts on a regional landscape is the result of many different factors. These include the nature and pattern of various fissures and vents, the length of time that volcanism is active, the relative age of volcanism, the composition and physical characteristics of extruded materials, the volume of erupted material, and the amount and extent of subsequent erosion. In some cases, voluminous lava flows and thick blanket tephras that accumulate over large areas may partially or entirely bury the preexisting landscape, whereas, in others, a focused distribution of lavas and tephras may produce a distinctive assemblage of lava-capped hills and mesas.
Full-text available
Pahoehoe and Aa basaltic lavas flows are common in volcano islands such as Hawaii, Reunion Island or Iceland. They are present in every large igneous provinces (LIPs) and giant lava flows identified in planetary volcanology. In terms of both areal coverage and total volume, pahoehoe flows dominate basaltic lavas in the suaerial and submarine environments on Earth. Several processes for pahoehoe lava flows emplacement have been identified, at different scales, from the extrusion of small lobes to the formation of giant flows. The dynamics of the geological fluids allow today to propose a new paradigm of implementation of pahoehoe lavas. The kinematics of the fluid mecanics is expressed through steamtube (immaterial) becoming lava tubes and lava tunnels (TTL). The lava tunnel, hitherto considered a speleological curiosity, has a central role in the dynamics of geological fluids. The anastomosing network of the TTLs is constitutive of the dynamics of placement of the pahoehoe both in open channel flow and pipe flow (inflation). The pahoehoe fluid is apprehended as a reactor in its own right. Seat of many physicochemical reactions (fractional crystallization, exsolution, segregation, coalescence, buoyancy, in particular), and characterized by various interfaces and zones of transitions between the different states of matter in constant transformation (number of phases, Newtonian fluid, Bingham complex fluid, Herschel-Bulkley). Various volcano-geomorphological witnesses, at various scales of observation, both superficial and underground, sign the mechanisms involved in the dynamics of implementation and allow its identification in current and ancient flows. The paradigm of implementation by TTLs is not only founder in pahoehoe but also in the constitution of LIPs.
Full-text available
Après une mise en perspective historique nous présentons un rapide état des connaissances en matière de tunnels de lave et de spéléothèmes volcaniques (lavacicles). Nous jetons les bases de la contribution de la volcanospéléologie aux géosciences. Enfin, nous présentons une dizaine de tunnels de lave islandais remarquables avant de conclure sur le développement de la volcanospéléologie et l’intérêt croissant suscité par les zones volcaniques et géothermales auprès des voyageurs, spéléologues et explorateurs.
Full-text available
From 1986 to 1997, the Pu'u 'O'o-Kupaianaha eruption of Kilauea produced a vast pahoehoe flow field fed by lava tubes that extended 10-12 km from vents on the volcano's east rift zone to the ocean. Within a kilometer of the vent, tubes were as much as 20 m high and 10-25 m wide. On steep slopes (4-10 ø) a little farther away from the vent, some tubes formed by roofing over of lava channels. Lava streams were typically 1-2 m deep flowing within a tube that here was typically 5 m high and 3 m wide. On the coastal plain (< 1 ø), tubes within inflated sheet flows were completely filled, typically 1-2 m high, and several tens of meters wide. Tubes develop as a flow's crust grows on the top, bottom, and sides of the tubes, restricting the size of the fluid core. The tubes start out with nearly elliptical cross-sectional shapes, many times wider than high. Broad, flat sheet flows evolve into elongate tumuli with an axial crack as the flanks of the original flow were progressively buried by breakouts. Temperature measurements and the presence of stalactites in active tubes confirmed that the tube walls were above the solidus and subject to melting. Sometimes, the tubes began downcutting. Progressive downcutting was frequently observed through skylights; a rate of 10 cm/d was measured at one skylight for nearly 2 months.
Full-text available
A revised regular solution-type thermodynamic model for twelve-component silicate liquids in the system SiO2-TiO2-Al2O3-Fe2O3-Cr2O3-FeO-MgO-CaO-Na2O-K2O-P2O5-H2O is calibrated. The model is referenced to previously published standard state thermodynamic properties and is derived from a set of internally consistent thermodynamic models for solid solutions of the igneous rock forming minerals, including: (Mg,Fe2+,Ca)-olivines, (Na,Mg,Fe2+,Ca)M2 (Mg,Fe2+, Ti, Fe3+, Al)M1 (Fe3+, Al,Si)2 TETO6-pyroxenes, (Na,Ca,K)-feldspars, (Mg,Fe2+) (Fe3+, Al, Cr)2O4-(Mg,Fe2+)2 TiO4 spinels and (Fe2+, Mg, Mn2+)TiO3-Fe2O3 rhombohedral oxides. The calibration utilizes over 2,500 experimentally determined compositions of silicate liquids coexisting at known temperatures, pressures and oxygen fugacities with apatite feldspar leucite olivine pyroxene quartz rhombohedral oxides spinel whitlockite water. The model is applicable to natural magmatic compositions (both hydrous and anhydrous), ranging from potash ankaratrites to rhyolites, over the temperature (T) range 900–1700C and pressures (P) up to 4 GPa. The model is implemented as a software package (MELTS) which may be used to simulate igneous processes such as (1) equilibrium or fractional crystallization, (2) isothermal, isenthalpic or isochoric assimilation, and (3) degassing of volatiles. Phase equilibria are predicted using the MELTS package by specifying bulk composition of the system and either (1) T and P, (2) enthalpy (H) and P, (3) entropy (S) and P, or (4) T and volume (V). Phase relations in systems open to oxygen are determined by directly specifying the f o 2 or the T-P-f o 2 (or equivalently H-P-f o 2, S-P-f o 2, T-V-f o 2) evolution path. Calculations are performed by constrained minimization of the appropriate thermodynamic potential. Compositions and proportions of solids and liquids in the equilibrium assemblage are computed.
Full-text available
Lava tubes play a pivotal role in the formation of many lava flow fields. A detailed examination of several compound `a`a lava flow fields on Etna confirmed that a complex network of tubes forms at successively higher levels within the flow field, and that tubes generally advance by processes that include flow inflation and tube coalescence. Flow inflation is commonly followed by the formation of major, first-order ephemeral vents which, in turn, form an arterial tube network. Tube coalescence occurs when lava breaks through the roof or wall of an older lava tube; this can result in the unexpected appearance of vents several kilometers downstream. A close examination of underground features allowed us to distinguish between ephemeral vent formation and tube coalescence, both of which are responsible for abrupt changes in level or flow direction of lava within tubes on Etna. Ephemeral vent formation on the surface is frequently recorded underground by a marked increase in size of the tube immediately upstream of these vents. When the lining of an inflated tube has collapsed, `a`a clinker is commonly seen in the roof and walls of the tube, and this is used to infer that inflation has taken place in the distal part of an `a`a lava flow. Tube coalescence is recognised either from the compound shape of tube sections, or from breached levees, lava falls, inclined grooves or other structures on the walls and roof. Our observations confirm the importance of lava tubes in the evolution of extensive pahoehoe and `a`a flow fields on Etna.
Lava tubes are a very common and important feature in mafic lava flows. The insulation provided by lava tubes allows molten lava to travel large distances from the vent with little cooling. This paper presentes the first attempt to quantify the processes that control this cooling. The resulting thermal budget balances heat loss by (1) conduction, (2) convection of air in the wall rocks, (3) vaporization of rainwater, and (4) radiation out of skylights against (1) viscous dissipation, (2) latent heat released during crystallization, and (3) the cooling of the lava. When applied to the Waha'ula tube on Kilauea Volcano, Hawaii, this thermal budget reproduces the observed ~1°C/km cooling of the lava inside the active tube. The thermal budget is also used to compare the insulating ability of hypothetical lava tubes in a continental flood basalt setting, on the ocean floor, Venus, the Moon, and Mars. This analysis demonstrates the large effect of rainfall and atmospheric convection, the importance of the volumetric flux of lava through the tube, and the overwhelming importance of compositional (i.e., rheological) differences. This work suggests that basaltic tube-fed flows several hundred kilometers long can be produced by eruptions with effusion rates of only a few tens of cubic meters per second. Thus even the longest lava flows observed in our solar system could have been produced by low to moderate effusion rate eruptions, if they were tube-fed. .
We describe the reactivation and the successive evolution of the shallow plumbing system of Mt. Etna between the end of the largest flank eruption of the last three centuries (1991–1993) and the subterminal eruption from South-East Crater (SEC), which occurred between February and mid-November 1999. Our analysis is based on observations of the volcanic activity and petrological studies of the erupted volcanics. Bulk rock, mineral and glass compositions have been determined for more than 80 samples erupted from the four summit craters between October 1995 and February 1999. These data allow us to recognise significant compositional variations among the products of different craters. In particular, volcanics produced between 1995 and 1999 by Bocca Nuova (BN), Voragine (VOR) and North-East Crater (NEC) show limited compositional variations and are similar to those observed during recent eruptions (e.g., 1991–93). More primitive magmas have been produced during the more vigorous fire fountains episodes. On the contrary, the South-East Crater produced slightly more differentiated volcanics than those of the other summit craters following its reactivation (November 1996) until the end of 1998. Whole rock compositions of products from this crater show low CaO/Al2O3, whereas interstitial glasses have lower MgO and higher alkali contents than those from the other craters. However, since the beginning of 1999, and just before the start of the subterminal eruption from SEC, the volcanics erupted from this crater progressively changed in composition, becoming similar to those of the other craters. This trend indicates that within the conduits of the summit craters, distinct thermal and fluid-dynamical regimes can evolve, controlling the cooling and crystallisation of Etna magmas.
1] Lava pillars are hollow, vertical chimneys of solid basaltic lava that are common features within the collapsed interiors of submarine sheet flows on intermediate and fast spreading mid-ocean ridges. They are morphologically similar to lava trees that form on land when lava overruns forested areas, but the sides of lava pillars are covered with distinctive, evenly spaced, thin, horizontal lava crusts, referred to hereafter as ''lava shelves.'' Lava stalactites up to 5 cm long on the undersides of these shelves are evidence that cavities filled with a hot vapor phase existed temporarily beneath each crust. During the submarine eruption of Axial Volcano in 1998 on the Juan de Fuca Ridge a monitoring instrument, called VSM2, became embedded in the upper crust of a lava flow that produced 3-to 5-m-high lava pillars. A pressure sensor in the instrument showed that the 1998 lobate sheet flow inflated 3.5 m and then drained out again in only 2.5 hours. These data provide the first quantitative constraints on the timescale of lava pillar formation and the rates of submarine lava flow inflation and drainback. They also allow comparisons to lava flow inflation rates observed on land, to theoretical models of crust formation on submarine lava, and to previous models of pillar formation. A new model is presented for the rhythmic formation of alternating lava crusts and vapor cavities to explain how stacks of lava shelves are formed on the sides of lava pillars during continuous lava drainback. Each vapor cavity is created between a stranded crust and the subsiding lava surface. A hot vapor phase forms within each cavity as seawater is syringed through tiny cracks in the stranded crust above. Eventually, the subsiding lava causes the crust above to fail, quenching the hot cavity and forming the next lava crust. During the 1998 eruption at Axial Volcano, this process repeated itself about every 2 min during the 81-min-long drainback phase of the eruption, based on the thickness and spacing of the lava shelves. The VSM2 data show that lava pillars are formed during short-lived eruptions in which inflation and drainback follow each other in rapid succession and that pillars record physical evidence that can be used to interpret the dynamics of seafloor eruptions. INDEX TERMS: 3035 Marine Geology and Geophysics: Midocean ridge processes; 3045 Marine Geology and Geophysics: Seafloor morphology and bottom photography; 3094 Marine Geology and Geophysics: Instruments and techniques; 8419 Volcanology: Eruption monitoring (7280); 8429 Volcanology: Lava rheology and morphology; KEYWORDS: submarine volcanic eruption, lava flow inflation, sheet flow, lava pillars Citation: Chadwick, W. W., Jr., Quantitative constraints on the growth of submarine lava pillars from a monitoring instrument that was caught in a lava flow, J. Geophys. Res., 108(B11), 2534, doi:10.1029/2003JB002422, 2003.
Explosive activity at Mt. Etna from July 19 to August 7, 2001, provides a good case study to investigate the causes of the transitions between style of basaltic explosion. In this period, a new vent, located at 2550 m above sea level on the southern flank of the volcano, exhibited three types of activities that followed one another: initial ash and steam explosions with the emission of radial jets, of hydromagmatic origin; intermediate fire fountaining and Strombolian explosions, due to magma vesiculation; and finally, sustained to pulsing ash explosions, caused by overpressurization of the degassed and cooling top of the magma column. The activities produced two end-members of juvenile ash in the size range 0.4–0.1 mm: (1) brown, fluidal- to irregular-shaped, vesicular sideromelane glass particles, and (2) microcrystalline, blocky, poorly vesicular tachylite particles. Component analysis of the ash reveals a gradual decrease in the abundance of sideromelane, replaced by tachylite, in the transition from the Strombolian to the final ash explosion activity. Dense blocks with irregular, variable surface textures also characterize the products of the late pulsing ash explosions. Petrographic, chemical, and crystal size distribution analyses, together with morphological evidences, indicate that sideromelane quenched earlier than tachylite during the final stage of magma evolution. In fact, the groundmass of tachylite formed by subsequent crystallization of magma, possibly at lower temperature and under different degassing conditions. We hypothesize that sideromelane formed in the central part of the volcanic conduit, where the buoyant rise of gas bubbles caused a higher magma ascent velocity, which did not allow time for vesicle to escape or collapse before fragmentation. Conversely, tachylite crystallized at the margins of the conduit, where slow-moving magma accumulated, temperature was lower, and vesicle collapsed, forming a network of cracks favorable to permeable gas flow. Reduced magma emission rate at the end of the Strombolian phase caused an increase in the thickness of the peripheral degassed magma zone, until it formed a plug at the top of the conduit, and activity gradually shifted to pulsing ash explosions. These were driven by repeated explosions of the overpressurized plug, in a small-scale, vulcanian-like, explosive process. We suggest that the relative abundance of sideromelane and tachylite ash particles in basaltic explosion products may provide information on the evolution of velocity gradients within magma flux.
Bibliogr. s. 355 - 362