Discovery of self-combusting volcanic sulfur flows
Andrew J.L. Harris*
Department of Earth Sciences, Open University, Milton Keynes MK7 6AA, UK
Sarah B. Sherman
Department of Geology and Geophysics, University of Hawaii, 2525 Correa Road, Honolulu, Hawaii 96822, USA
Robert Wright*
Department of Earth Sciences, Open University, Milton Keynes MK7 6AA, UK
ABSTRACT
Hitherto sulfur flows have been recognized as lobate features similar in form to basaltic lava
flows. However, we have discovered a self-combusting sulfur-flow mode that leaves an entirely
different and unexpected deposit. In this mode, the flow is emplaced in a combusting state, so that
all sulfur is burned away to leave a sulfur-free, thermally eroded trough. During the 4-hr-long
event we observed, combustion of 0.6 m3 of sulfur generated 2.4 tons of SO2. Once under way,
combusting flows do not require eruption of a molten volume to maintain activity: They can
generate supply volume by melting surficial sulfur along the flow path. Combusted flow features
are widespread at Vulcano, Italy, indicating that this previously unknown emplacement mode
may be common. Previous failure to recognize this flow style may account for the apparent rarity
of sulfur flows. Our new findings overturn conventional thought on how sulfur flows are emplaced,
interpreted, and considered, and may show them to be a common volcanic feature.
Keywords:Vulcano volcano, sulfur flow, flow emplacement, combustion, sulfur dioxide, thermal erosion.
INTRODUCTION
During October 1998 we witnessed the emplacement of a combusting sulfur flow at Vulcano
volcano, Italy, a phenomenon that has never previously been documented. Moreover, our observations reveal a completely different flow behavior to that previously observed, showing that a
relatively common feature has been misunderstood and, as a result, overlooked. Our findings
have huge implications for the manner in which
sulfur flows are examined and considered. These
observations are surprising because we overturn
conventional wisdom where, traditionally, sulfur
flows have been envisaged as lobate green-yellow
features that have proved difficult to find. As a
consequence, sulfur flows are considered rare. We
show that sulfur flows can, due to the combusting
emplacement mechanism described herein, look
very different and leave instead a gray-encrusted,
rubble-lined trench. Such features are common on
Vulcano, suggesting that sulfur flows are not as
rare as previously thought. Given that combustion
may be common and cause frequent, but temporary, increases in the sulfur dioxide flux, combusting flow events may play an important role in
the production of this volcanogenic gas.
Noncombusted sulfur flow, lake, and pyroclastic deposits are relatively well documented,
having been described on Azufre (Galapagos),
Mauna Loa (Hawaii), Lastarria (Chile), and Poas
(Costa Rica) (Banfield, 1954; Colony and Nordlie,
1973; Skinner, 1970; Greeley et al., 1984;
Naranjo, 1985; Oppenheimer, 1992; Oppenheimer
and Stevenson, 1989; Francis et al., 1980). The
case described here, however, is only the second
time that observations of an active sulfur flow have
been reported; the previous time was on SiretokoIosan volcano (Japan) in 1936 (Wantanabe, 1940).
Our observations differ significantly from those of
Wantanabe (1940), providing a rare insight into
sulfur-flow emplacement. The Siretoko-Iosan
flows were characterized by torrents of molten
sulfur that solidified to leave pahoehoe- and ‘a‘alike features. The Vulcano flow, however, was emplaced in a burning state, where combustion left a
sulfur-free, thermally eroded trench largely devoid
of typical flow features.
To date, the Siretoko-Iosan noncombusting
emplacement mode and deposit have been used as
the sole criteria for postemplacement sulfur-flow
identification (Banfield, 1954; Colony and Nordlie,
1973; Greeley et al., 1984; Naranjo, 1985; Skinner,
1970) and for volcanic flow-emplacement simulations (Greeley et al., 1990). Considerations of
sulfur-flow occurrence and processes are therefore incomplete. Key flow features and localities
may thus have been overlooked on this planet, as
well as on Io, where noncombusted terrestrial
analogues have been used to guide feature interpretation (e.g., Carr et al., 1979; Sagan, 1979).
Here we describe the new self-combusting emplacement mode and deposit type, permitting a
new consideration and inventory of a previously
poorly understood geologic feature.
*Present addresses: Hawaii Institute of Geophysics
and Planetology, School of Ocean and Earth Science
and Technology, University of Hawaii, 2525 Correa
Road, Honolulu, Hawaii 96822, USA.
TEMPERATURE-DEPENDENT
BEHAVIOR OF SULFUR
Two temperature-dependent properties of
liquid sulfur (color and viscosity) are important in
Geology; May 2000; v. 28; no. 5; p. 415–418; 3 figures.
sulfur-flow observations. Yellow α- and β-sulfur
alotropes melt at 112.8–115.1 °C and 114.6–
119.6 °C, respectively (Meyer, 1976). With increased temperature, color changes from yellow
to orange (~133 °C), red-orange (~161 °C), red
(~171 °C), and brown-black (~221 °C), before
reaching its boiling point of 444.6 °C (Meyer,
1976; Pieri et al., 1984). At 159.4 °C liquid sulfur undergoes a sudden increase in viscosity
(Meyer, 1976) from ~10 –2 to ~10 2 Pa s, transforming a low-viscosity, fast-moving flow into a
high-viscosity, slow-moving flow.
FIELD SETTING: THE VULCANO
FUMAROLE FIELD
Since the last magmatic eruption of Vulcano
during 1888–1890, activity has been characterized by fumarolic activity, currently concentrated in a 16 300 m2 zone extending 260 m
along the northeast crater rim and 140 m from
the rim to the floor. During October 1998 we
measured maximum and mean vent temperatures of 446 °C and 169 ± 72 °C at this fumarole
field (1213 measurements). Our thermal survey
showed that the field was dominated by lowtemperature fumaroles, with 84% of the vents
having temperatures of <250 °C, and 33% at
<120 °C. Below ~120 °C, fumaroles are encrusted with yellow sulfur. Such low-temperature fumaroles are typically clustered in 50–
300-cm-diameter, 1–10-cm-deep pits lined with
trachytic-rhyolitic and sulfur blocks and coated
with centimeter-thick powdery yellow-orange
sulfur (Fig. 1) containing rare beads and driblets
of molten, orange sulfur. The flow occurred
within a low-temperature fumarole zone that
extends craterward from the rim in the eastern
415
(A)
(B)
F
p
b
s
s
s
s
s
s
s
1m
Sulfur lake and flow complex
Sulfur-rich zone
s
Sulfur-poor zone
Residual trachyte block
s
Late-stage sulfur flow
Sulfur-crusted fumarole
Location of 96 °C fumarole
Figure 1. A: Features of combusting sulfur flow observed during emplacement. Main photograph
shows flow extending on 2 branches on either side of 35-cm-long hammer. Red inset (upper
right): primary lake (bounded by dashed black lines) detailing very light gray ash crust. Green
inset (middle right): low-temperature fumarole pit within which primary lake developed. Blue
inset (lower right): molten, very dark red and moderate olive-brown early-stage flows. Yellow
inset (lower left): ash-crusted mature flows (bounded by dashed black line). Flow-marginal
deposits of very light gray crust reveal earlier, higher flow levels. Magenta inset (middle left):
sulfur block during melting; crusted secondary lakes are forming at its base. B: Features of combusting sulfur flow observed following emplacement, where tape is 86 cm long, flow direction is
top to bottom (indicated by red arrows in red inset), and dashed black lines show former lake
and/or flow extents. Main photograph shows combusted flow section (postevent). Blue inset
(lower right): rubble-lined trench and rare late-stage flows. Red inset (upper right): typical lake
deposit with coating of very light gray ash and trachytic blocks that have collapsed inward owing
to removal of support. Green inset (middle right): typical (rubble-lined trench) flow deposit.
sector of the field, across which maximum and
mean temperatures in October 1998 were 289 °C
and 131 ± 44 °C, respectively. The zone is characterized by discontinuous coverage of lowtemperature fumarole pits, separated by ground
covered by trachytic-rhyolitic blocks set in a
case-hardened ash matrix, with common sulfurous rinds (Fig. 1A).
OBSERVATIONS OF A SELFCOMBUSTING SULFUR FLOW
Our observations allowed us to map features
characteristic of this combusting sulfur flow
(Fig. 2) and describe the five-stage evolution
that led to these features (Fig. 3). We define
flow color using the Geological Society of
America rock-color chart (Rock Color Chart
Committee, 1991) and use this color terminology herein.
Stage 1: Primary Lake Establishment
The flow source was a 24 × 20 cm, 5–10-cmdeep sulfur lake that developed at about 14:00
(all times are local) on October 4, 1998 (Fig.
416
3A). The lake was established immediately
downslope of a 55-cm-wide, low-temperature
(≤157 °C) fumarole pit, across which normal
activity persisted throughout our observations.
From the sulfur encrustations evident around the
lake, we assume that the lake basin was originally a similar low-temperature fumarole pit.
The lake core was very dark red, had a temperature of 120–150 °C, and was effervescing vigorously. Blue flames played at the surface, giving
temperatures of 270–470 °C at a 20 cm height.
Initially, plastic sulfur crusts developed, which
quickly became covered by very light gray ash
(Fig. 1A). This evolution was typical of all lakes
and flows observed, and we assume that it was
due to the mixing of trachytic-rhyolitic ash with
the molten sulfur, the ash becoming concentrated at the surface by burning off the sulfur
component. Sulfur blocks and encrustations next
to the lake were melted to contribute to the lake
volume, such that lake filling and overflow resulted in flow generation (Fig. 1A). A second
cause of flow onset was by melting a breach in
the sulfur-block retaining walls.
Figure 2. Sulfur flow map. Flow directions are
given by arrows. F—unaffected fumarole pit,
p—primary lake, s—secondary lakes, and b—
trachytic block displaying sulfur-rind burn
(checkerboard shows burn zone).
Stage 2: Flow Extension
Flows fed by breaching the primary lake wall
were yellow and moderate olive-brown with temperatures of <120 °C (Fig. 3A). These formed
flow fields of 2–5-cm-wide pahoehoe lobes on
fronts ≤15 cm across. Within the flow field, lower
viscosity, higher temperature (120–160 °C) very
dark red sulfur flowed in 1–3-cm-wide leveed
channels, which became deepened to depths of
1–2 cm by thermal erosion of underlying flow
sulfur and sulfur in the sulfur-trachyte-rhyolite
substrate. As at the primary lake, channelized
molten sulfur effervesced and burned with blue
flames. Within 10–15 min, the initial channels
and pahoehoe flows were overrun and melted by
faster flowing, lower viscosity flows of burning
very dark red and moderate olive-brown sulfur
(Fig. 1A). These flowed in torrents over the preexisting flows, developed plastic crusts at temperatures of 250–300 °C, and quickly destroyed all
of the initial yellow pahoehoe and leveed features.
Stage 3: Secondary Lake and Flow Formation
About nine low-temperature fumarole pits
were present downslope from the primary lake.
On extending into these pits, the burning flow
melted any sulfur encrustations and blocks
present. In this way, secondary lakes were established that had the same characteristics as, and
were directly linked to, the primary lake (Fig. 3B).
Secondary lakes were generated by explosive
ejection of burning sulfur cinders from active
GEOLOGY, May 2000
Figure 3. Schematic plots showA
B (B)
G
H (D)
L
K
(A)
(C)
ing combusted flow evolution.
1
13
B
11
4
p
p
G p H 1
K
2
L
A
A: Establishment of primary lake
D
E
I J
C
14
(p) and flow extension (black)
s
s
s F
f
s
s
f
toward fumarole pits (f). B: Sec15
2
12
3
ondary lake formation by flow exs
s
f
f
Sulfur
s
s
f
f
tension into sulfur pits and sulfur
rich zone
N
s
s
f
f
M
cinder ejection (s). C: Establishs
s
f
f
ment of mature lake and flow
C
D
I
J
complex, where older lakes and
1
s
7
F
E
f
M
N s
5 f
11
flows (gray tones) are beginning
8
17 1
to mature. D: Lake and flow stagnation where activity has died
Sulfur poor zone
15
2
out owing to supply exhaustion
2
16
12
6
3
10
9
3
in proximal parts (dots), activity
is stagnating in medial to distal
parts, and single flow unit extending into sulfur-poor zone has stalled. Lake and flow cross sections summarize characteristic features typical of
each stage; 1—flame, 2—molten lake and/or flow, 3—effervescence, 4—plastic crust, 5—levee, 6—thermal erosion, 7—pahoehoe-lobe emplacement, 8—cinder ejection, 9 and 10—rootless secondary lake generation and sulfur-rind burn-off due to cinder impact, respectively, 11—extensive
ash crust, 12—flow and/or lake level decline, 13—remnant overhang and crust deposit, 14—fumarole reestablishment, 15—remnant trachyte
block, 16—burning, stagnant sulfur lake, and 17—trench.
flows or lakes. On landing in fumarole pits,
cinders melted any sulfur contained therein to
generate secondary lakes with no obvious, direct
connection to the primary lake. Cinders landing
on fumarole-free surfaces had insufficient source
material to generate new lakes or flows; however,
any sulfurous rinds were burned off to leave a
blackened scar. Noncombustion of explosively
ejected sulfur deposited spatter as far as 2–3 m
from the flow. Such explosive ejection could be
responsible for sulfur spatter and spray observed
around flows at Mauna Loa and Siretoko-Iosan
(Greeley et al., 1984; Wantanabe, 1940).
Stage 4: Mature Lake and Flow Complex
Establishment
After ~1 hr, primary lake activity began to
decline. Continued sulfur melting and combustion resulted in steady supply exhaustion and
consequent lake-level decreases (Fig. 3C). The
same process occurred at all lakes and flows,
causing flows to reach maturity <1 hr after initial
emplacement. As at the lakes, continued combustion resulted in lowering of flow levels and
formation of extensive very light gray ash crusts
covering a molten interior of very dark red sulfur.
With further combustion and flow-level decline,
coatings and overhangs of gray crust marked
higher flow stands (Fig. 1A).
Stage 5: Lake and Flow Stagnation
By 15:00, the lake and flow complex had
extended to the lower edge of the fumarole field.
Flows from the final lake in the complex then
spread into a zone free of fumarole pits and
dominated by trachytic-rhyolitic blocks. Unable
to generate any further supply by melting sulfur
in its path, this distal flow extended ~1 m into the
sulfur-free zone before outrunning its supply and
stagnating (Fig. 2). By 16:00, activity at the primary lake had ceased: all available sulfur had
been melted and combusted. Activity continued
at secondary lakes and flows, but lake levels were
declining, and flows were stagnating (Fig. 3D).
GEOLOGY, May 2000
In this final phase, melting of partially melted
sulfur blocks continued at the still-active lakes,
but melt supply was no longer sufficient to feed
flow. Forward flow motion ceased, and a string of
stagnant, very dark red sulfur pools with extensive very light gray ash crusts formed along the
former flow path. Combustion continued at all
lakes and flows, so that by 18:00, activity at most
locations had ceased because of supply exhaustion: All available sulfur had been melted and
burned off. By this time, a 95 °C fumarole had
been reestablished at the primary lake location.
REMNANT FEATURES AND EVENT
FREQUENCY
Melting and combustion excavated a sinuous
15–20-cm-wide, 10–15-cm-deep trench encrusted
with very light gray ash deposits and lined with
trachyte-rhyolite blocks. Lake sites left roughly
oval, deeper pits within, and sometimes wider
than, the trench to contribute a beaded form to the
trench (Fig. 1B). Rare yellow-green sulfur-flow
deposits with lobes and channels remained, but
these were late-stage features, emplaced after the
main stages of flow emplacement and burning had
ceased. Late-stage features were typically 30 cm
long and 15 cm wide, compared with the ~700 ×
25–125 cm dimensions of the combusted flow
field, occupying <5% of the field area (Fig. 1B).
Having recognized combusted sulfur-flow
features and deposits, it became apparent that this
emplacement mode was spatially common at
Vulcano. Previously recognized centimeterscale, yellow lobate flows were reassessed as
minor, late-stage components of larger, meterscale, combusted flows whose deposits they lay
within. Cooler fumarole field areas contain ubiquitous combusted sulfur-flow features, the largest
of which was a 10-m-long, 1–2-m-wide, <0.5-mdeep thermal-erosion channel that fed a 13 × 8 m
thermally eroded zone of dispersed flow. The
channel displayed overhanging rims, studded
with protruding trachyte-rhyolite fragments, and
traces of ashy residue. Numerous thermally
eroded lake basins were discovered, including an
~0.4-m-deep cylindrical hole, feeding an ~0.5-mwide, 0.35-m-deep channel that tunneled under a
2-m-wide trachyte-rhyolite block. That ephemeral combusting-sulfur events may also be temporally common was indicated by observation of
a second event at Vulcano in 1995, when we discovered a 20 × 10 cm, ~10-cm-deep burning
sulfur lake in a low-temperature fumarole zone at
the crater rim. Although a flow was not observed
from this lake, it displayed characteristics identical to those of the 1998 event.
Localized combustion may occur on noncombusting flows. The smooth, overhanging southern
margin of the Mauna Loa flow described by
Greeley et al. (1984) and assumed by them to be
the result of ablation-erosion processes shows
characteristics identical to those of partially
melted sulfur blocks at Vulcano. We examined a
sulfur block taken from the Mauna Loa flow.
Although this showed pahoehoe-type structures,
it also possessed smooth, concave surfaces identical to those at partially melted blocks at Vulcano.
Where melting is partial, melt-generated features
(e.g., smooth, concave surfaces) will mark the
melting limits, and identification of postemplacement sulfur-flow areas by virtue of pahoehoe- and
‘a‘a-type features will underestimate the flow
extent. Thus previously examined flows may have
been more extensive than originally thought.
SULFUR-FLOW COMBUSTION: SULFUR
DIOXIDE GENERATION AND CAUSE
Sulfur combustion temperature in air at standard pressure is 248–261 °C; sulfur will burn
with a blue flame to yield sulfur dioxide: Sn +
nO2 → nSO2 (Meyer, 1977). In the observed
case, combustion converted >95% of the sulfur in
the flow path to SO2 . The reaction resulted in
SO2 atmospheric concentrations of >50 ppm (our
gas-mask-filter tolerance limit). Effervescence
occurred along the flow length and was always
associated with combustion. We therefore assume
that effervescence resulted from flow-wide con417
version of S to SO2 . Given the dimensions of
thermally eroded channels measured by us, we
estimate that a typical Vulcano sulfur flow will
melt 0.6–44 m3 or 1.2–88 tons of S. This equates
to 2.4–176 tons of SO2 that, if liberated over the
observed 4-hr-long activity period, would generate a SO2 flux of 0.6–44 tons/hr. This compares
with measured SO2 fluxes at Vulcano of 1.7–5
tons/hr (Bruno et al., 1993). At Vulcano, SO2 flux
during periods of combusting flow activity will
be on the order of that due to magma degassing.
Coincidence of SO2 measurements with a combusting flow event could therefore result in a
doubling of measured fluxes. During such an
event, because SO2 liberated during combustion
of surface deposits represents the release of
stored S, assuming that the total flux is due to
magma degassing would cause erroneous conclusions regarding degassing rates.
To begin the melting and combustion process,
temperatures must rise from below the sulfur
melting point to the combustion point. During the
observed event, fumarole temperatures near the
flow remained low and stable; <158 °C in adjacent, unaffected, pits. Across the fumarole field,
maximum and mean temperatures were lower
than in previous years (we obtained 516–541 °C
and 202–216 °C, respectively, during 1994 –
1995). The required temperature increase cannot
therefore be explained by increased field or zonal
temperatures (n.b.: this would also have triggered
flows at other fumaroles). The cause must therefore have been localized to the fumarole at which
the flow began. We propose that the trigger was
the failure of a sulfur cap that had developed to
seal the affected fumarole prior to the flow event.
During 1998, we analyzed sulfur precipitation
using quartz glass tubes buried, and left for
~96 hr, in 6 fumaroles at temperatures of 95–
350 °C. Results showed that, once temperature
falls below 100 °C, conduit sealing by sulfur
precipitation proceeds at 0.009 kg · cm–1 · day –1
to build a massive, impermeable, crystalline cap.
Such seals are widespread at Vulcano, forming
sulfur veins along fissures and within fumarole
pits. Cap failure would cause the sudden release
of high-temperature, pressurized gas trapped in
the conduit below the seal. This gas would carry
the heat required to begin melting and combustion. By providing a proximal sulfur source, the
seal would contribute a molten volume to initiate
primary lake formation. In support of this model,
zones of extensive flow coverage coincide with
areas where sealed caps are common.
DISCUSSION AND CONCLUSIONS
At previously documented sulfur flows, chambers have been assumed to evolve during subsurface temperature increases, the sulfur then being
erupted at the surface (Naranjo, 1985; Oppenheimer, 1992; Oppenheimer and Stevenson,
1989; Skinner, 1970; Wantanabe, 1940). In our
case, supply was maintained by flow extension
418
into surficial sulfur deposits, where melting due
to the heat of combustion generated a new supply
volume that was incorporated into the flow volume. This required no subsurface heating, caused
the flow to be self-perpetuating, and meant that
the flow length was limited by the areal extent of
surface sulfur deposits, rather than by the volume
of subsurface melt.
Sulfur volcanism plays a fundamental role on
Io, a moon of Jupiter (Sagan, 1979; Pieri et al.,
1984). The unique sulfur-flow emplacement style
and remnant features observed have implications
for the interpretation of Io data. Although high
temperatures calculated for Io indicate that silicate volcanism may dominate (e.g., Lopes-Gautier
et al., 1997; McEwen et al., 1998), extension of
silicate flows (or any material at a temperature
higher than the sulfur melting point) into sulfurrich deposits will generate secondary sulfur
flows. Once under way, such flows will be capable
of sustaining flow by melting sulfur in their path.
Eruption of sulfur is therefore unnecessary for
flow generation, and flows can propagate as long
as a supply of surface sulfur is available. By melting the sulfur component of the substrate,
thermal erosion channels will be among the features symptomatic of such activity.
Two end-member modes of sulfur-flow emplacement exist (noncombusting and combusting), each resulting in markedly different deposits.
Until now, sulfur flows have been identified by
sulfur pahoehoe lobes, ‘a‘a, leveed channels, and
tubes. However, flow combustion will strip sulfur
from all invaded surfaces so that most features
conventionally used to identify sulfur flows will
not survive. At Vulcano, features symptomatic of
combusting emplacement are more common
than their noncombusted counterparts. Thus field
evidence such as sulfur-free thermal-erosion
channels should be used in sulfur-flow mapping.
At Vulcano, such considerations transform an
apparently rare event into one that is spatially,
and probably temporally, common.
ACKNOWLEDGMENTS
We are grateful to D. Pirie and R. Carniel for field
assistance, and to P. Wallace, S. Hiroshi, D. Pieri, D.
Rothery, and M. Widdowson for reviews. Harris was
funded by the Open University Research Development Fund and a Leverhulme Trust Special Research
Fellowship.
Carr, M.H., Masursky, H., Strom, R.G., and Terrile,
R.J., 1979, Volcanic features on Io: Nature,
v. 280, p. 729–733.
Colony, W.E., and Nordlie, B.E., 1973, Liquid sulfur at
Volcan Azufre, Galapagos Islands: Economic
Geology, v. 68, p. 371–380.
Francis, P.W., Thorpe, R.S., Brown, G.C., and Glasscock, J., 1980, Pyroclastic sulphur eruption at Poas
volcano, Costa Rica: Nature, v. 283, p. 754–756.
Greeley, R., Theilig, E., and Christensen, P., 1984, The
Mauna Loa sulfur flow as an analog to secondary
sulfur flows(?) on Io: Icarus, v. 60, p. 189–199.
Greeley, R., Lee, S.W., Crown, D.A., and Lancaster, N.,
1990, Observations of industrial sulfur flows:
Implications for Io: Icarus, v. 84, p. 374–402.
Lopes-Gautier, R., Davies, A.G., Carlson, R., Smythe,
W., Kamp, L., Soderblom, L., Leader, F.E.,
Mehlman, R., and the Galileo NIMS Team, 1997,
Hot spots on Io: Initial results from Galileo’s near
infrared mapping spectrometer: Geophysical Research Letters, v. 24, p. 2439–2442.
McEwen, A.S., Keszthelyi, L., Spencer, J.R., Schubert,
G., Matson, D.L., Lopes-Gautier, R., Klaasen,
K.P., Johnson, T.V., Head, J.W., Geissler, P.,
Fagents, S., Davies, A.G., Carr, M.H., Breneman,
H.H., and Belton, M.J.S., 1998, High-temperature silicate volcanism on Jupiter’s moon Io: Science, v. 281, p. 87–90.
Meyer, B., 1976, Elemental sulfur: Chemical Reviews,
v. 76, p. 367–388.
Meyer, C., and Meyer, B., 1977, Sulfur, energy, and
environment: Amsterdam, Elsevier, p. 38–79.
Naranjo, J.A., 1985, Sulphur flows at Lastarria volcano
in the north Chilean Andes: Nature, v. 313,
p. 778–780.
Oppenheimer, C., 1992, Sulphur eruptions at Volcano
Poas, Costa Rica: Journal of Volcanology and
Geothermal Research, v. 49, p. 1–21.
Oppenheimer, C., and Stevenson, D.S., 1989, Liquid
sulphur lakes at Poas volcano: Nature, v. 342,
p. 790–793.
Pieri, D.C., Baloga, S.M., Nelson, R.M., and Sagan, C.,
1984, Sulfur flows of Ra Patera, Io: Icarus, v. 60,
p. 685–700.
Sagan, C., 1979, Sulphur flows on Io: Nature, v. 280,
p. 750–753.
Skinner, B.J., 1970, A sulfur lava flow on Mauna Loa:
Pacific Science, v. 24, p. 144–145.
Rock Color Chart Committee, 1991, The Geological
Society of America rock color chart: Boulder,
Colorado, Geological Society of America, 8 p.
Wantanabe, T., 1940, Eruptions of molten sulphur from
the Siretoko-Iosan Volcano, Hokkaido, Japan:
Japanese Journal of Geology and Geography,
v. 17, p. 289–310.
Manuscript received October 25, 1999
Revised manuscript received January 31, 2000
Manuscript accepted February 9, 2000
REFERENCES CITED
Banfield, A.F., 1954, Volcanic deposits of elemental
sulfur: Canadian Mining and Metallurgical Bulletin, v. 47, p. 769–775.
Bruno, N., Caltabiano, T., Grasso, M.F., Porto, M., and
Romano, R., 1993, SO2 flux COSPEC measurement: Acta Vulcanologica, v. 3, p. 269.
Printed in USA
GEOLOGY, May 2000