Pele`s hair and tears-their origin and composition
Pele`s hair and tears-their origin and composition
Pele`s hair and tears-their
origin and composition
Diana Oettel (48943), TU Bergakademie Freiberg

Fig.1. Pele-the goddess of fire and volcanoes (WEB 1)
Abstract. Lava Fountains are one of the most impressive volcanic
phenomena. In these fountains lava is ejected to great heights under high
speeds and temperatures, producing pyroclastic products such as lapilli,
spatter, pumice and scoria. This paper presents a collection of chemical and
physical studies on Pele´s hair and tears, showing their association according
to their formation. Chemical zonation rims on both of them allow to assume an
interaction of these glasses with acid gases in the plume. Furthermore vesicle
studies display volatile exsolutions before and after the eruption.
Introduction

Fig.2.
Pele`s hair (WEB 2) chased her to the next island. After a

Fig.3.
Pele`s tears (WEB 3)
fragments of glass. Limu o Pele is
formed when water is forced into and trapped inside lava,
as when waves wash over the top of the exposed flows of the molten rock. While
boiling the water is converted to steam, expanding to form bubbles within the
lava. The lava rapidly cools and the glass bubbles burst[2].
However this paper is dealing with two other forms occuring also due to the
effect of rapid cooling: Pele`s hair and tears, which are glassy materials
formed during fountaining of fludial lava in the air[3].
The origin of Pele´s hair and tears
Oftentimes Pele is shown in pictures in connection with the
violent interaction of water, lava, and explosive eruptions. But typically
Hawaiian eruptions are the calmest of the eruption types. The type of eruption
and also the type of volcano depend on the gas content, the magma temperature
and the composition of the magma. Hawaiian magmas are among the hottest on
earth. Temperatures of approximately 2.200°F (1204°C) have been measured in
molten lava (Kilauea, Hawaii)[4].
Therefore Hawaiian eruptions are characterized by gentle emission of very fluid
lava with less gas contents. Compared with other eruptions, the hawaiian type
is less explosive and the relative volume of ejected pyroclastic material is
lower. The gas escaping at the vent forms huge clouds[5].
The most impressiv attribut of hawaiian-style eruptions are their high (tens-to
hundreds of meters) lava fountains, which shoot into the air, while magma rises
up. Clots, which rise within the fountain, are generally still hot and
incandescent when the fall back to the surface and can accumulate to form lava
flows. These lava flows are about 30 feet (9,14 m) thick. Long series of
eruptions can build masses of lava flows downslope around the vent. Small hills
are called lava shield and large, broad ones are called shield volcanoes[6].

Fig.4. Shield volcano (HAZLETT,
1996)
In Figure 4 you can see a cut through a typical Hawaiian
shield volcano. While shield volcanoes stand above hot spots they grow rapidly
and their mounds can enlarge to have tens of miles in diameter.
Observations show that lava fountains
occur in short spurts or can last for hours. Under certain conditions scoria
and spatter is formed and accumulated downward. The basalt can quench into a
recitulite (glassy rock) during periods of high vesiculation. But the smallest
pyroclasts carried downward are Pele´s tears. During high winds they drawn out
to form long filaments called Pele´s hair[7].
Natural Glass
Basaltic glasses are of interest as they are the major constituents
of many products (e.g. Pele`s seaweed and recitulite) from basaltic volcanism. Minerals
in general have regular geometric arrays. But glasses originates from magma
with such a high cooling rate (shown in the Figure 5), that the atoms in the
melt can not organize in a crystal lattice[8].

Fig.5.
Cooling rate of Pele´s hair (GÖTZE, 2008)
Instead it forms a viscous amorphous
glass, which can be explained as a solution of elements. These elements are for
example Al, Si, O, Ca, K. Because glass is formed under special conditions it
is metastabile near the surface, and Calcium and potassium get lost first.
Furthermore the glass starts to recrystallize into more stable minerals. Thats
the reason why most of the glasses are only of the cenozoic age. As shown in
Figure 5 different kinds of natural glass are formed under special conditions.
Pele´s hair generates by cooling down from a very high temperature with the
highest cooling rate. Other glasses produced by some lightning strikes
(Fulgurite), by frictional processes in fault zones, burning of underground
coal and impact of large meteorites have either lower temperatures or cool down
more slowly. Fast chilling of magma during building of basaltic ash cones can
result in the formation of the brown glass sideromelane[9].
Furthermore rhyolithic glass (> 69% SiO2) is particularly
widespread because of the high viscosity of silica melt. An example is
Obsidian. While rhyolitic glasses are grey-black, the increasing content of Ca,
Mg (decreasing Si and K) produce cinnamon-brown colors in thin sections of
basaltic glasses[10].
Pele`s hair is especially of interest to show the element composition of
hawaiian tholeiitic melts. Two rock types represent either a rock composition
mixture of Pele`s hair and olivine or a composition of parental magma, which
seems to be the result of subtracting plagioclase and clinopyroxene from the
magma of Pele`s hair11.
Pele´s hair and tears
Morphology
Pele´s tears are spherical pyroclasts, which have sizes
varying from few µm to hundreds of µm of diameter. As shown in Figure 6 (part A
of SEI) their surfaces are rough and uneven. The prevalent bumpiness and their
appearance on all sides of the tears suggests particle-particle interaction. In
case of shocking during ground impact events, one or more deformed sides would
appear in a none-spheri-
cal size. As shown by the arrows in
Figure 6, small particles adhere to the surface of the Pele's tear. They are
interpreted as sublimates condensed from gases of the plume or small glass
particles12. The hair, shown in Figure 7 has a cylindrical
form, is 8 mm long and has a width varying between 1-500 µm in diameter. Pele´s
hair can be up to 2 m long, but is never found complete, because of breaking
during transport in the wind or contact with the ground. The hair is associated
with a “knot” in the middle and a droplet at the end. If you assume that hairs
can be formed from stretching into strands, than knots are maybe the result of
not stretchable crystals enclosed in the glass. If you have a closer look at
this hair, you can see oftentimes entire and broken vesicles, which have
parallel orientation to the axis of elongation. Sometimes you can find
particles adhering to Pele`s hair, too and these particles may have distinct
chemical characteristics. In this case, they have a high cloride content, which
is typical for the Masaya gas plume.13


Fig.6. Pele´s tear from Masaya Volcano (secondary electron image,
JOEL 5910LV: accelerating voltage-15kV, probe current - 2 nA, working distance
- 19 mm; MOUNE et. al., 2007)

Fig.7. Pele`s hair from Masaya Volcano
(secondary electron image, JOEL 5910LV: accelerating voltage-15kV, probe
current - 2 nA, working distance - 19 mm; MOUNE et. al., 2007)
12 HEIKEN
and WOHLETZ, 1985, page 245
13 MOUNE et.
al., 2007, page 245
Formation under certain conditions
Some experiments on ink jets produzed from nozzle were
performed using different Weber number (We)
and Reynolds number (Re). As shown in
the following Figure, if Re is large
in comparison with We a droplet is
produced. On the other hand if We is
large and Re is low a thread is the
result.
Shimozuru defined Pele`s number (Pe) as the ratio of (We~Re).

Pe = We~Re = pv η/ po σ (3)
p is the density of the liquid v is the velocity of
liquid flow l is the effective length of flowing liquid σ is the surface tension of the liquid. po and η denote the density and viscosity of
the liquid
Fig.8. Ink jet produced from nozzle under
different We and Re-time sequence from top to bottom (Left:We = 20, Re = 44; Right: We =
50, Re =32); SHIMOZURU,
1994
Considering that p/
po
in equation 3, velocity, surface tension and viscosity of the liquid flow are
the parameters of remainder for observation. The temperature of the lava is
involved in vicosity and surface tension. Measurements of the viscosity of
hawaiian melts show results of 87 Pa s and 48 Pa s at temperatures of 1160°C
and 1190°C (documented temperatures of Alae eruption at Kilauea). So if the
range in temperature is small, a small range in viscosity is the result. The
effects of temperature of magma on surface tension of the liquid are different.
Positive temperature derivates[11]
and surface tension decreases with increasing temperatures[12]
were determined. But at least surface tension has also not high effects on
Pele´s number. Thus, the most significant parameter is eruption velocity. To
conclude: If the velocity is high- Pele´s number became large and Pele´s hair
will be produced. On the contrary Pele`s tears are formed.
Experiments with artificial glass fibre
also display low viscosities (<100 Pa s) spurting of molten
silicate. To produce a high quality fibre glass a certain viscosity of the
silicate and the temperature of re-heating, the combination of flow velocity
from the nozzle and also revolving speed of the spinner[13]
is necessary. That means Pele´s hair might be produced under very complicated
conditions including turbulences while spurting.
Transport and Interaction

Fig.9.
Cavity of Pele`s hair (MOUNE et.
al., 2007) pothesis
says that cavities act like fun-
nels and samplers of tears during trans-
port of Pele´s hair in the volcanic plume. Another argument
to support this theory is that Pele´s tears are observed in high concentrations
at the bottom of the cavities. So it is possible that Pele´s hair and tears are
associated due to the effects of transport and independent of their mechanism
of formation. The arrows in Figure 9 show that sublimates also adhere to the
surface of the tears in the cavities, as shown in Figure 6 (page 6).

these edges form a kind of barrier during transport,
so that the tears can trap.
Remarkable is that these tears have a high varity in their
sizes. In Backscat-
|
Fig.10.
Cross-section of Pele`s tear (MOUNE et. al., 2007)
|
tered Electron Images (BSEI) of Pele´s
tears spherical gas bubbles can be found (Figure 10). Dark
material in the bubbles is derived from polishing. The bubbles can be up to 150
µm in diameter in tears with sizes of 800 µm). Furthermore a tabular shaped
crystal of plagioclase is linked to the gas bubble. Remarkable is also the
chemical zonation rim of this droplet[14],
which is between 6 and 10 µm broad.
Chemical Composition

Fig.11. The
total alkali-silica (TAS) diagram showing three analysis of Pele´s hair and one
of Pele´s tear (modified after Streckeisen et. al., 1985, complete chemical
data see Appendix)
Figure 11 displays three analysis of Pele`s hair and one of
Pele`s tear, which are plotted in the TAS-diagram. They all have a basaltic
composition and you can see that Pele`s tear (the point in the top right corner
of the Basalt field) has a different

Fig.12.
Chemical zonation rim (MOUNE shown in Figure 12). Furthermore mea-
et. al., 2007)
surements of the outer part of the tears
were also taken. A strong chemical gradient in
concentration of the major elements was measured comparing the outer part and
the inner part of the tears. At a distance of 2 µm and 6 µm a much higher SiO2
content was observed. The average of analysis represent lower SiO2 content
and small variation in the inner part of the tear[15].
The major element totals are normalized at 100%. In
the outer zone of the tear the major element totals are too low. In contrast,
the totals of the analysis in the inner part of the tear are much higher
(97–98%) and homogeneous (Figure 13). The sums of the analyses 1 and 2 (86 to
88 wt.%) performed in the outer zone are maybe a result of the high volatile
content in the tears suggesting that the
Masaya melt was not totally degassed while erupting. The chemical variability
from the tears' interior to its rim display increasing silica enrichment and
decreases when all other elements increase.

Fig.13. Compositional profile along the line of
measurements
(MOUNE et. al., 2007)
Pele´s hairs also contain euhedral plagioclase
crystals and chemical zonation around the external part. Chemical zonation can
also be found in cavities on their internal walls along Pele´s hairs. The
thickness of silica enrichment at the cavity walls inside and outside is
constant. Some tears inside these cavities display chemical zonation, some have
no zonation and others are transformed.
Furthermore along-axis sections of Pele´s hair show
spherical and curved sizes, so that they are not always elongated to the hair
axis. This variations can be observed on the same hair[16].
So a complex formation of Pele´s hair can be assumed (compared with page 7).
Diana Oettel
Conclusion
Pele´s hair and tears are fundamentally composed of glass,
and that infers a rapid cooling in the eruptive plume. These pyroclastic
materials are unique, caused by their formation including rapid quenching and
exceptional high temperatures of magma resulting from fountaining of
hawaiian-style eruptions. The fact that cavities in Pele´s hairs act like
funnels of Pele´s tears displays that they are associated due to the transport
and formation. The most important parameter, which decides whether a droplet or
a hair is formed is the spurting velocity. If it is high Pele´s hair is formed,
it not Pele´s tears are the result.
The glass of Pele´s tears also contains
euhedral plagioclase crystals. That shows that the magma was not super-heated
and cooling was rapid as these crystals do not expose dentritic overgrows. The
plume temperature was probable below the glass transition temperature. This is
a supposition made because of the absence of devitrification textures. Vesicles
in tears and hairs suggest that the magma was not totally degassed at the
eruption-time. Pele`s tears seems to have a general concentration of volatiles
between 2.9-1,9 wt.% [17].
Another community of Pele´s hair and
tears is the chemical zonation rim. This rim is best explained by dissolution
of silica glass during interaction with volcanic gases in the plume and
potentially a important chronometer of residence time in the plume[18].
Alteration by rain water is not likely because samples were collected
immediately after eruption. Furthermore the Silica enrichment is not connected
with enrichments in Al and Ti, which should be the effect of different mobility
of network forming and modifying cations[19].
After all the scenario of forming Pele´s hair and tears can be decribed as the
following:
At the beginning of the eruption the
spurting velocity is high during fountaining at the top of the conduit. Pele´s
hair is formed. Vesicles obsevered in Pele´s hair have parallel orientation to
the axis of elongation of the hair. This is a result, if these vesicles are
deformed with regard to direction of magmatic gas jets. Later, when the spurting
velocity is decreased (the fountain might be less high), but the temperature is
still high enough, that Pele`s hair stays in liquid state, gas exolution
produces sperical vesicles. Furthermore Pele`s tears are produced at this time,
bacause they contains only spherical vesicles. No stretching is associated with
the formation of Pele´s tears. This result is consistent with the model, saying
that Pele`s tears are produced when spurting velocity is low. Before the
ejection, there are several turbulent mortions inside of the eruptive plume at
relatively low temperatures included in this model. Thats a suggestion based on
the fact that no devitrification textures are observed[20].
Pele´s hair and tears-their origin and composition
References
BEST,
Myron G.(2001); Christiansen, Eric H.: Igneous Petrology. Blackwell Science,
Inc. Press: page 22
CAMP,Vic(2006),
Department of Geological Sciences, San Diego State University, link:-www.geology.sdsu.edu/how_volcanoes_work/Hawaiian.html
CLAGUE,
D. A.; Davis A.S.; Bischoff, J. L.; Dixon, J. E.; Geyer, R.(2000): Lava
bubble-wall fragments formed by submarine hydrovolcanic explosions on
Loìhi Seamount and Kilauea Volcano. Springer Verlag, Bull
Volcano
61:437-449: page 438
DE BOER, Jelle
Zeilinga; Sanders, Donald Theodore(2005): Volcanoes in Human
History. Princeton University Press: page 34
DEVINE, J.D.,
Sigurdsson, H., Davis, A.N., Self, S.(1984): Estimates of sulfur and chlorine
yield to the atmosphere from volcanic eruptions and potential climatic effects.
Journal of Geophysical Research 89, 6309– 6325.
GÖTZE
J.(2008/2009), Angewandte Mineralogie, Skript 6 link:-www.mineral.tu-freiberg.de/mineralogie/mintech/lehre/lehrmateri-
HAZLETT,
Richard W.; Hyndman, Donald W.(1996): Roadside Geology of Hawaii. Mountain
Press Publishing Company: page 13
HEIKEN, G., Wohletz
K.(1985): Volcanic Ash, University of California press,
Berkeley, California: page 245
KATSURA, Takashi
(1967): Pele´s hair as a liquid of Hawaiian tholeiitic basalts.
Geochemical Journal, Vol. 1: pages 157 to 168
KING,
T.B.(1951): The surface tension of silicate slugs. J Soc Glass Tech 35:
241-259.
MOUNE, Séverine;
Faure, François; Gauthier, Pierre-J.; Sims, Kenneth W.W.(2007): Pele`s hairs
and tears: Natural probe of volcanic plume. Journal of Volcanology and
Geothermal Research, Vol. 164 issue 4: pages 244 to 253.
MURASE
T.; McBirney A.R.(1973): Properties of some common igneous rocks and their
melts. Geol Soc Am Bull 84: 3563-3592.
Diana Oettel
PARFITT, Elisabeth
A.(1998): A study of clast size distribution, ash deposition and fragmentation
in a Hawaiian-style volcanic eruption. Journal of Volcanology and Geothermal
Research 84: pages 197 to 208 link:-www.sciencedirekt.com
POTUZAK, M.;
Dingwell D.B.; Nichols A.R.I.(2006): Hyperqueched Subarial
Pele`s hair Glasses from Kilauea Volcano, Hawaii.
Geophysical Research Abstracts, Vol.8, 06908: page 1
SHIMOZURU,
D.(1994): Physical parameters governing the formation of Pele's hair and tears,
Bulletin of Volcanology 56: pages 217 to 219 link:-www.springerlink.com
STERPENICH, J.,
Libourel, G.(2001):Using stained glass windows to understand the durability of
toxic waste matrices. Chemical Geology 174: pages 181–193.
List of Figures
Figure 1: Pele- goddess of fire and volcanoes
Figure 2: Pele´s hair
Figure 3: Pele´s tears
Figure 4: Shield volcano
HAZLETT, 1996. page 13
Figure 5: Cooling rate of Pele´s hair
GÖTZE, 2008
Figure 6: Pele´s tear from Masaya Volcano
Figure 7: Pele`s hair from Masaya Volcano
6, 7:MOUNE et. al., 2007, page 246
Figure 8: Ink
jet produced from nozzle
SHIMOZURU, 1994, pages 217 to 219
Figure 8: Cavity of Pele`s hair
Figure 9: Cross-section of Pele´s tear
Figure 10: Chemical zonation rim
8, 9, 10: MOUNE
et. al., 2007, pages 247-248
Figure 11: The total alkali-silica (TAS) diagram showing three
analysis of Pele`s hair and one of Pele´s tear modified after Streckeisen et. al., 1985
Figure 12: Chemical zonation rim
Figure 13: Compositional profile along the line of measurements
12, 13: MOUNE et. al., 2007, pages 248-249
Pele´s hair and tears-their origin and composition
Appendix
|
Pele`s tears
|
Pele`s hair
|
|
|
|
|
1
|
2
|
3
|
SiO
2
TiO
2
A O
l2 3
FeO
MnO
MgO
CaO Na O
2
K O
2
|
50,9 (0,6)
1,42 (0,13)
13,5 (0,4)
13,8 (0,4)
0,25 (0,08)
4,67 (0,14) 8,81 (0,30)
2,83 (0,18)
1,39 (0,13)
|
48,82
2,77
13,42
9,9
0,18
9
11,32
2,25
0,58
|
50,26
2,69
13,48
9,57
0,17
7,04
11,45
2,22
0,45
|
50,04
3,02
14,02
9,45
0,17
6,93
11,45
2,42
0,57
|
Total
|
97,6 (0,8)
|
100,22
|
99,96
|
100,24
|
Pele´s tears: Masaya
Volcano (Nicaragua)
MOUNE, Séverine;
Faure, François; Gauthier, Pierre-J.; Sims, Kenneth W.W.(2007): Pele`s hairs
and tears: Natural probe of volcanic plume. Journal of Volcanology and
Geothermal Research, Vol. 164 issue 4: pages 244 to 253.
Pele´s hairs: Kilauea
(Hawaii)
KATSURA, Takashi
(1967): Pele´s hair as a liquid of Hawaiian tholeiitic basalts. Geochemical
Journal, Vol. 1: pages 157 to 168
[1]
DE BOER, 2005, page 34
[2]
CLAQUE et. al., 1998, page 438
[3]
POTUZAK et. al., 2006, page 1
[4] HAZLETT and HYNDMAN, 1996,
page 13
[5] MACDONALD et.al., 1983,
page 9
[6] PARFITT, 1998, page 197
[7] CAMP, 2006
[11]
MURASE and MCBIRNEY, 1973
[12]
KING, 1951
[13]
SHIMOZURU, 1994, page 218
[14] MOUNE et. al., 2007, page 246
[15]
MOUNE et. al., 2007, page 247
[16] MOUNE et. al., 2007, page
249
[17]
DEVINE et. al., 1984
[18]
MOUNE et. al., 2007, page 250
[19]
STERPENICH and LIBOURAL, 2001, pages 181-193
[20]
MOUNE et. al., 2007, page 251