Journal of Volcanology and Geothermal Research 197 (2010) 188–208

Contents lists available at ScienceDirect

Journal of Volcanology and Geothermal Research

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j vo l g e o r e s

The 1793 eruption of San Martín Tuxtla volcano, Veracruz, Mexico

J.M. Espíndola a,⁎, A. Zamora-Camacho a, M.L. Godinez b, P. Schaaf c, S.R. Rodríguez d
a
Departamento de Vulcanología, Instituto de Geofísica, Universidad Nacional Autónoma de México, Coyoacán 04510 D.F., Mexico
b
Departamento de Geografía Económica, Instituto de Geografía, Universidad Nacional Autónoma de México, Coyoacán 04510 D.F., Mexico
c
Departamento de Geomagnetismo y Exploración, Instituto de Geofísica, Universidad Nacional Autónoma de México, Coyoacán 04510 D.F., Mexico
d
Departamento de Geología Regional, Instituto de Geología, Universidad Nacional Autónoma de México, Coyoacán 04510 D.F., Mexico

a r t i c l e i n f o a b s t r a c t

Article history: San Martín Tuxtla (N18.562°; W95.199°, 1659 masl) is a basaltic volcano located in southern Veracruz, a
Received 23 October 2008 Mexican State bordering the Gulf of Mexico. It rises in a volcanic field strewn with monogenetic volcanic
Accepted 11 August 2009 cones, maars and three other large volcanoes mostly dormant since the late Pliocene: Santa Marta, San
Available online 22 August 2009
Martín Pajapan and Cerro El Vigía. The latest eruptive event of San Martín occurred in 1793 and was
described by Don José Mariano Moziño, a naturalist under the commission of the Viceroy of the then New
Keywords:
Spain. In this work we present results of the study of this eruption based on historical accounts and field
San Martín Tuxtla volcano
basaltic volcanoes
observations. We identified an ash deposit around the volcano related to the 1793 eruption, mapped its
Mexican volcanoes distribution and determined its granulometric, petrographic and geochemical characteristics. These studies
explosive basaltic eruptions suggest that the volcano began its activity with explosive phreatomagmatic explosions, which were followed
alkaline volcanism by Strombolian activity; this period lasting from March to October 1793. The activity continued with an
effusive phase that lasted probably 2 years. The eruption covered an area of about 480 km2 with at least 1 cm
of ash; the fines reaching distances greater than 300 km from the crater. A total mass of about 2.5 × 1014 g
was ejected and the volcanic columns probably reached altitudes of the order of 10 km during the most
explosive phases. The lava emitted formed a coulee that descended the northern flank of the volcano and has
an approximate volume of 2.0 × 107 m3.
© 2009 Elsevier B.V. All rights reserved.

1. Introduction devoted to the first two aspects of the ECP, but few to the volcanological
aspects of the recent volcanism in the TVF. Previous studies of the TVF
San Martín volcano (N18.562°; W95.199°, 1659 masl) is one of the have been devoted to the geochemical characteristics of its rocks.
many Mexican volcanoes that have been active during historical times Friedlaender and Sonder (1923), in one of the earliest studies of the TVF,
(Fig. 1). It is the highest peak within the Tuxtla Volcanic Field (TVF), also described the general characteristics of the area and reported the
known as Los Tuxtlas Volcanic Field or Los Tuxtlas Massif, which consists earliest petrographic and geochemical analyses of its main volcanoes.
of hundreds of scoria cones, maars, and three additional large volcanic Although they considered that the rocks were of calc-alkaline nature, a
edifices mostly dormant since late Pliocene: Santa Marta, San Martín re-examination of the data by Pichler and Weyl (1976) placed them in
Pajapan, and Cerro el Vigía. The large physiographic area on which the the alkaline suite in agreement with further analyses of other rocks from
TVF rises, the eastern coastal plains (ECP), has attracted the attention of the area. Thorpe (1977) also noted the alkaline nature of these rocks
many researchers for reasons that can be summarized as follows: First, after analyzing samples from several scoria cones. Aguilera Gómez
the area constitutes the W and SW portions of the Gulf of Mexico margin (1988) obtained ages from different rocks of the TVF and established
and therefore provides clues on its formation. Second, some of its that the oldest stem from the Pliocene. Nelson and González-Caver
geological units are important oil and gas bearers. Third, the (1992) and Nelson et al. (1995) carried out a more comprehensive study
manifestation of Cenozoic alkaline volcanism in different parts of the on the geochemistry and chronology of the rocks of the area finding a
ECP poses important questions regarding the origin of such magmatism predominance of alkaline rocks and concluding that the activity
and its relationship with the tectonic environment. Finally, the occurred in two periods separated by a hiatus of quiescence. Thus,
occurrence of historical volcanic activity in the TVF calls for attention they divided the rocks from the volcanic field into the older and younger
to potential volcanic hazards in the area. Several studies have been volcanic series (OVS and YVS respectively). The rocks from the OVS
range in age from 7 to 2.6 Ma and underlie the YVS. They crop out at the
SE and W portions of the TVF and include the large edifices of the Santa
Marta, San Martín Pajapan and Cerro el Vigía volcanoes (Fig. 2). The
⁎ Corresponding author. rocks from the YVS were extruded from 0.8 Ma to the present and
E-mail address: jmec@unam.mx (J.M. Espíndola). include San Martín volcano and a large number of scoria cones and

0377-0273/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jvolgeores.2009.08.005
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 189

Fig. 1. Location of the Tuxtla Volcanic Field in the State of Veracruz, Mexico. MBV: Mexican Volcanic Belt. Notice the low topography along the margin of the Gulf of Mexico and the
isolated position of the Tuxtla Volcanic Field with respect to the Mexican Volcanic Belt.

maars around its edifice. Because no rocks showed ages in the range 2.6 for these two ages the eruptive history of San Martín volcano is
to 0.8 Ma, Nelson and González-Caver (1992) suggested that this was a essentially unknown. Such a large structure must have been
period of quiescence. Reinhardt (1991) carried out a geomorphologic constructed through many eruptive episodes, varying in style and
study of the cinder cones of the YVS and concluded that they are intensity, as evidenced by the morphology of the volcano. The edifice
younger than 50,000 years; he also found that to the west of San Andrés of San Martín is a massive structure with a crater slightly less than a
Tuxtla and north of Catemaco at least 9 eruptions took place within the kilometer in diameter. The rim of the crater is somewhat irregular in
last 6000 years. These eruptions are related to scoria cones and maars perimeter and very irregular in height, with altitudes ranging mostly
located in that area and the chronology was established from between 1500 and 1600 masl. The highest peak is on the east side and
radiocarbon dating of charcoal samples and archeological data derived reaches 1659 masl. The crater has steep inner walls and bears two
from the nearby Matacapan archeological site, together with strati- cinder cones in its interior; at least one of them formed during the last
graphic correlations. Nelson and González-Caver (1992) and Nelson eruption of San Martín, as will be described below (Fig. 3). The
et al. (1995) considered that magmatism of the TVF had signatures of eruptive history of the volcano is complex, and the unraveling of the
subduction related to the Cocos plate. Verma (2006), however, on the stratigraphic record represents an arduous task since most deposits
basis of further geochemical analyses of rocks of the area, argued that are considerably altered due to their basaltic composition and the
magmas at the TVF have no significant contribution from the downgoing action of the tropical climate. In addition, San Martín deposits are
Cocos plate and that they are similar to those found in rift or extensional intercalated with deposits from hundreds of Pleistocene–Holocene
settings. More comprehensive investigations, including geophysical scoria cones and maars, and covered extensively by vegetation.
studies, will be required to resolve the origin of the TVF magmatism. However, the volcano has erupted in historical times, yielding at least
The activity of San Martín volcano has been addressed only one documented example of its activity. Fortunately, the products of
marginally in previous studies. Thus, this paper is focused on the this latest eruption can still be observed directly at several sites
activity of this volcano and in particular on the eruption of 1793. We around the volcano.
reconstructed the characteristics of the eruptive period through the The latest eruption of this volcano occurred when the present
analysis of historical accounts, field observations, and granulometric territory of Mexico was part of Spain's viceroyalty of New Spain.
and geochemical analyses. The results of this study are relevant to Medel y Alvarado (1963), chronicler of the nearby town of San
understanding basaltic volcanism in general and provide context to Andrés, makes reference to the existence of Ixtlán, a village formerly
evaluate volcanic hazards in this area. located on the western slopes of the volcano, which early after the
conquest of Mexico by the Spaniards, was covered by ash from an
2. Volcanic activity at San Martín Tuxtla volcano eruption of the volcano occurring probably in 1534. According to the
chronicler this eruption compelled the inhabitants to abandon their
The oldest age known of a rock from San Martín Tuxtla volcano village and settle farther away from the volcano, a migration that led
(SMTV) is the 0.8 Ma obtained by Nelson and Gonzalez-Caver (1992) to the foundation of Tuxtla (now Santiago Tuxtla, Fig. 2) and Tzacoalco
for a basanite lava sample collected from its base. These lavas are the (now San Andrés Tuxtla, Fig. 2). It is doubtful whether an eruption
oldest rocks that can be related to an eruption from the volcano. A occurred in that year because there are no other accounts or proofs of
layer of highly altered ash is fairly well distributed around the volcano its occurrence; unfortunately Medel y Alvarado does not provide the
and overlies a paleosol with an age of 4500 years BP (Table 1). Except source of this information. A later eruption, which occurred in 1664, is
 190
J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

Fig. 2. Geology of the Tuxtla Volcanic Field (modified after Nelson and González Caver, 1992).
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 191

Table 1
Conventional radiocarbon ages for samples collected around San Martín volcano, Tuxtla Volcanic Field, Veracruz, Mexico.

Section Sample numbera Type Age (BP) δ13

SM1-16 A-14234 Burned log under lava flow 290 ± 40 − 27.0

SM2-45C A-12646 Brown paleosol under dark gray ash (1793) 290 ± 45 − 25.6
SM9-1b A-11107 Brown paleosol under dark gray ash (1793) 360 ± 50 − 25.4
SM5-14c A-14237 Brown paleosol under dark gray ash (1793) 400 ± 60 − 25.2
SM-13b A-13512 Brown paleosol under dark gray ash (1793) 405 ± 45 − 25.5
SM1-5b A-12643 Brown paleosol under dark gray ash (1793) 490 ± 45 − 25.2
SM4-17c A-13504 Charcoal shards in brown paleosol under dark gray ash (1793) 490 ± 125 − 29.2
SM3-15a A-13508 Brown paleosol under dark gray ash (1793) 535 ± 55 − 16.6
SM4-7b A-13510 Brown paleosol under dark gray ash (1793) 555 ± 45 − 25.6
SM5-8a A-14236 Brown paleosol under dark gray ash (1793) 570 ± 45 − 26.2
SM0-2 A-11106 Brown paleosol under dark gray ash (1793) 615 ± 55 − 25.3
SM1-4a A-13505 Light brown paleosol under altered ochre and gray lapilli and ash beds 1060 ± 40 − 24.6
SM2-15c A-12645 Light brown paleosol under altered ochre and reddish lapilli and ash beds 1475 ± 45 − 25.0
SM2-10c A-13507 Light brown paleosol under altered ochre and whitish lapilli beds 2700 ± 40 − 9.2
SM2-15a A-12644 Brown paleosol under altered ochre and reddish lapilli beds 4080 ± 55 − 23.3
SM0-18b A-11109 Brown paleosol under altered reddish scoria and over lava flow 5390 ± 55 − 25.2
SM2-7b A-14235 Paleosol under altered ochre lapilli bed 27,750 + 795 − 22.8
− 720
a
Dating was carried out at the Laboratory of Isotope Geochemistry, Department of Geosciences, The University of Arizona, Tucson, Arizona.

mentioned in two sources although with no agreement as to the eruption. Moziño was a naturalist born and educated in the New Spain
month and day. The first and oldest of these sources are the report well known for his botanical work with the Royal Scientific Expedition
written by D. Joseph Mariano Moziño on occasion of the 1793 to New Spain (Engstrand, 1981). After the explosive events that

Fig. 3. Topographic model of the edifice of San Martín showing the inner cones, the lava flow and surrounding area (Topographic data from: Hoja E15A63 Nueva Victoria, Instituto
Nacional de Estadistica y Geografía, Mexico. Scale 1:50,000; the processing was carried out with ®ArcGIS).
192 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

Table 2
A summary of the main events of the 1793 eruption of San Martín Tuxtla volcano (Veracruz, Mexico) as extracted from the account of Don José Moziño (1870).

Date Event Comment

2 March; 16:00 h First eruption Thunder and sight of a column of fire, ash falls. Lasted 2 days.
Phreatomagmatic (?)
22 May; 7:00 A.M. Second eruption Larger than the first one higher column and larger amounts of
Phreatomagmatic (?) ash dispersed. Lasted 2 days.
23 May Ash reaches the city of Oaxaca, other towns and the Lasted at least 2 days.
province of Tabasco
28 June; 6:00 A.M. Third and largest eruption: heavy ash falls and thunder. Lasted 3 days.
Phreatomagmatic (?)
26 August at night Fourth eruption: ash falls and heavy rainfalls Lasted about two months.
23 September First ascent of D. José Moziño and companions
21 November Second ascent of D. José Moziño and his companions Artist D. Antonio Echeverría accompanies the exploring party
who witnessed another eruption

initiated the 1793 eruptive period of San Martín, and given the conducted his investigation. Inquiring into the previous behavior of
generalized panic that ensued among the inhabitants of the the volcano, he reported that:
neighboring regions, Moziño was sent by the Count of Revillagigedo,
Viceroy of New Spain to report on the eruption. A copy of the report according to the accounts from the elders of this neighborhood, last
for the Viceroy is conserved at the “Archivo General de Indias” (AGI) century the mountain of San Martín, located some 2 leagues north
in Seville, Spain. It was also published in 1870 by the “Sociedad from the village of San Andrés ejected flames and sand: they assure
Mexicana de Geografía y Estadistica” (SMGE) in its Bulletin (Moziño, that this happened on a 15 of October, with no memory of the year or
1870). Moziño's report is a somewhat general account of the eruption any other testimony, except for a recollection made by the
of 1793 but also refers to an older eruption that he placed in the 17th commanding officer of the troops, a truthful and old neighbor, of
century. From this report one derives that he arrived by boat from having read a property title that incidentally mentions a pledge made
Veracruz and stayed at today's San Andrés Tuxtla whence he to celebrate a religious feast dedicated to the glorious Spanish virgin

Fig. 4. Location of the main towns mentioned in the report on the 1793 eruption of San Martín Tuxtla volcano by Moziño (1870).
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 193

Santa Teresa de Jesus, to commemorate the event… the explosion From these two accounts it seems clear that, even allowing for the
then lasted a short time: the cast matter reached no more than 3 or 4 effect of time in people's memory, this eruption was of small duration
leagues, and left only smoke, which everybody looked at with little and without lasting consequences. The 1793 eruption, however, was
concern, even contempt, and which disappeared completely more of such magnitude that it induced the Viceroy to send one of the best
than 50 years ago. (Translated from Moziño (1870), p. 63). naturalists at his command to investigate the event. Thanks to this
decision we have a description of the eruption. In this paper we focus
in this eruption in an attempt to determine some of its characteristics.

The second source is a note published in 1830 by José Aurelio 3. The 1793 eruption of San Martín Tuxtla according to Moziño
García in El Constitucional, a newspaper published in Xalapa (capital
of the State of Veracruz) and later translated to German and published The report of Moziño on the 1793 event, which he partially
in the Neues Jahrbuch für Mineralogie (hereafter referred to as García, witnessed, has been the source of all references to this eruption in the
1835). We were unable to locate the original version in Spanish of this works of the early accounts on Mexican volcanism (Mühlenpfordt,
note and relied on the German translation. In this article García stated 1844; Pieschel, 1855). In a direct or indirect form it has also been the
that in 1824, when he was mayor of San Andrés, while paging source of later works such as those by Felix and Lenk (1890), Waitz
documents from the archives of the village, he came upon documents (1915), Sapper (1917), Friedlaender and Sonder (1923), and of
that described an eruption that took place on 15 January 1664. The several catalogues of volcanic activity such as Lamb's catalogue of
day of the date does not agree with the day gathered by Moziño (15 dust-producing volcanic eruptions (Lamb, 1970), IAVCEI's Catalogue
October), but the year is in agreement with the statements of the of Volcanoes of the World (Mooser et al., 1958), and the Smithsonian's
elders of the region as mentioned above. According to García, one Volcanoes of the World (Simkin and Siebert, 1994). In the latter the
morning, without notice, the sun suddenly darkened followed by a eruption is listed with a VEI = 4 and the emission of a volume of
rain of ash and sand accompanied by a rumbling noise coming from 108 m3 of tephra.
the mountain. People in San Andrés being very frightened sought Table 2 summarizes the most important aspects of the eruption as
refuge in the church, from where they were led in procession by the reported by Moziño. The eruption began on 2 March 1793 at 16:00 h;
priest to the head parish in Santiago Tuxtla where they remained however, since heavy clouds covered the mountain and thunder-
amid imploring chants until the sky cleared up. storms were frequent in its vicinity, most of the people took it for a

Fig. 5. Painting of the 1973 San Martín eruption by Don Antonio Echeverría. The legend in Spanish at the bottom of the painting translates as: “Crater of Tuztla's volcano seen from the
east in 1793. AB Diameter of the column of fire on 23 September. C Site where the first bottle with an inscription was buried. D Site where the second bottle with another inscription
was placed”. (Watercolor on canvas, copy courtesy of the Archivo General de Indias, Seville, Spain).
194 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

large thunderstorm. A couple of hours later, however, the cloud cover According to Moziño the 23 May fallout reached the towns of
disappeared, driven by the winds, and people were able to see: Oaxaca, Izucar, Tehuacán, Orizaba, Córdoba, and the province of
Tabasco. The most important towns of this province were Villaher-
a large column of fire… towards the northeast of Tuxtla and the
mosa, present capital of the State of Tabasco, and Tacotalpa, the capital
north of San Andrés; many thunderbolts stemming from its center
at that time, sited 45 km SSE from Villahermosa (Fig. 4). All of the
and swerving in different directions frightening the residents to
towns mentioned above are located at different azimuths, and
the point of rushing crowds to the temples to implore for divine
distances between about 230 and 320 km from the volcano. Moziño
mercy thinking that, if not the ruin of the universe, at least of the
estimated the area covered by ash at 11,000 squared Castilian leagues,
region was inevitable. (Translated from Moziño (1870), p. 63).
which is about 3.42 × 105 km2. This is an overestimation by a factor of
about 3 of the area within the boundaries marked by the referred
The eruption lasted for about two days with fluctuating intensity towns and cities, which is of about 1 × 105 km2. In order to cover this
and was characterized by abundant ash fallout towards the north due area with ash the volcanic cloud must have changed in direction
to the winds from the south. On 3 March, the explosions were heard at throughout the eruption, since these cities are located in a 180° sector
villages such as Perote, Teziutlan, Jalacingo, Papantla, Misantla and around the western side of the volcano.
Andrés Chalchicomula (now Ciudad Serdán), towns located at A third event occurred at 6:00 A.M. of 28 June. This was the largest
distances as far as 300 km from the volcano, and taken as cannonade eruption and the southern winds carried the ash towards the sea and
against pirates at the port of Veracruz. More than 400 “cannon shots” the hills of Tecolapan (NW from the volcano). As a result the roads in
were counted at Papantla and Misantla. that area were so much disfigured and made inaccessible by the ash
The second eruption, which lasted for about two days, occurred on cover that the postman had to return to Alvarado (a town 65 km NNW
22 May at 7:00 A.M. as the wind was blowing from the north: from the volcano) with his weekly mailbag. Woods were set aflame
during the 3 days that the eruption lasted.
The height of the fire column was greater than the first time, more
Finally, on 26 August, after a few weeks of quiescence, a new
frequent the lightning, more widespread the cloud of smoke and
period of activity ensued. This time ash fallout and lightning were
heavier the ash fallout. The sun was so much darkened that in a
accompanied by thunderstorms, which were frequent during the
radius of more than 15 leagues it was necessary to use artificial
rainy season. The eruptive episode lasted until an undetermined day
lighting at noontime. (Translated from Moziño (1870), p. 64).
of October, unknown because the heavy cloud cover surrounding the
mountains impeded the view.
Since Moziño used Castilian units of measurement, it is convenient at Moziño had arrived in the area in early September witnessing the
this point to introduce the equivalence with the modern metric system: continuous ash falls:
the Castilian foot, yard and league (pie, vara y liga castellanas) were
units of length commonly used during the XVIII century in Spain and its Since the beginning of September, being at sea because I was
territories. The foot was equivalent to 0.2786 m, 3 feet comprised a vara navigating towards this coast, I could feel the rain of sand that did not
or 0.8358 m, and a league comprised 20,000 feet or 5572 m. stop on all the days that followed during the trip along the rivers

Fig. 6. Topographic profile across San Martín volcano and the interior cones.
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 195

flowing towards Alvarado, and from those days to the present there poor condition of the instrument, which had been broken during the
has been hardly a day in which there has not been some large or trip from Mexico City, and was cursorily repaired in town. The
small fallout. (Translated from Moziño (1870), p. 65). naturalist found the volcano in another stage of activity, this time
greatly subdued, with a column reduced in height and width by about
2/3 and the emission of lava that slowly flowed to the north:
Later, on 23 September he ascended to the vent and estimated the
The pile of boulders seen to the north is threatening with collapse
width of the column in 40 yards and its height in more than 100 yards
because it has already beginning to topple over; the earthquakes
or 84 m and 334 m respectively. These coarse estimates correspond to
are stronger in that part, and there is a fire at the base, which
the period of relative weakening of the activity that allowed the
though not so fierce as before, nonetheless keeps digging the
ascent to the crater of the naturalist and his companions. He also
ground. (Translated from Moziño (1870), p. 68).
estimated that, in a circle of about 2 leagues of diameter around the
mountain, the fallout was at least 3 yards thick reaching up to 6 yards
in some places (see Appendix A). Based upon this data, he estimated Echeverría represented the eruption at the time of their visit in a
the minimum volume of the deposits in little more than 57 × 106 cubic painting conserved also in the AGI (Fig. 5). The painting illustrates the
feet, which is not in agreement with the above data and may be due to eastern side cone in its late stage of activity as seen from the east. It
an error in the transcription of his report. The data mentioned would depicts a small lava fountain stemming from a spot on the NW corner
yield a figure of 1.13 × 1010 cubic feet, a volume two orders of of the cone's crater; a deposit of smoking lava is seen on the ground to
magnitude higher. In metric units the deposit was estimated by the right of the cone. This eruption is listed in several of the works
Moziño at 2.5 to 5.0 m thick, the diameter of the circle at 5572 m and cited as the event of 21 November, although Moziño did not report on
therefore the corresponding volume of ash at 2.4 × 108 m3. its beginning and ending. Mooser et al. (1958) list an event on 23
He ascended for the second time on 21 November with a party that November, but it probably refers to this eruption.
included Don Antonio Echeverria, scientific illustrator of the Royal Felix and Lenk (1890) and Sapper (1917) after them, stated that
Expedition. On that occasion, from barometric measurements, he the activity continued through December of 1793 and that another
calculated the height of San Martín at 500 yards asl (418 masl). explosion, in which a small crater was created in May 1794. The
Compared with the actual elevations of the crater rim, this is a poor source of this information is Pieschel (1855), but this author cited as
estimate of the height of the volcano, which was probably due to the his source the report of Moziño, who did not mention these particular

Fig. 7. Outcrops of 1793 ash deposits at different sites around the volcano.
196 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

facts. Regardless of whether this explosion occurred or not it seems the aerial photograph where the apron of the eastern cone is seen
that the activity decreased gradually, probably with further bursts. resting over the margin of the western cone (WEC). This cone is also
According to García (1835) ash falls continued in the vicinity of the the one that produced the lava flow descending the northern side of
crater for two more years and “flames” could be observed preferably the volcano. However, Moziño visited the crater in September, almost
after stormy weather. In 1829, when he and some companions at the end of the active period, and it is not clear if he observed the
climbed to the summit, the cone was still smoking and the ground was western smallest cone. In Echeverria's painting the view of the WEC is
hot. The reference to ash falls occurring close to the crater for two obstructed by the EAC. It is therefore not known whether this cone
years suggests that lava emission continued also during that time was active during the first period of the 1793 eruption or if it was built
interval. during an older eruption. Moziño, however, refers to the EAC as the
Moziño ended his report commenting on the effects of the “mouth now burning” and he makes reference to the “inner rim of the
eruption on people, who were still afraid of a possible renewal of northwestern forge”, which could be a reference to the WEC. The lava
the activity, and on the subsequent harvests of corn, beans and cotton, flow stemming from the EAC extends for about 3 km from the rim of
which not only resulted unimpaired but also had an excellent yield. the crater and is composed of lava blocks covered by soil and dense
At present, the volcano shows no signs of further thermal activity, vegetation, which obscures its morphology (Fig. 3). A tree log at the
and the two cones are completely covered by a dense vegetation base of the flow (site SM1-16) yielded an age of 290 ± 40 years before
cover. In fact, there are no thermal springs around the volcano, even present (Table 1).
though there are many springs, which support all the water needs of
people and cattle at the base of the volcano, including the main town 4.2. Volume of cones and lava flow
of San Andrés Tuxtla. There is, however, a thermal anomaly on a spot
of the southern crater rim: at a place known to the inhabitants on the The EAC must have grown with time since, according to García
slopes of the volcano as “El caliente” (“the hot one”, Fig. 3); just a few (1835), the activity continued for several years. Given the amount of
cm below the surface and the soil, one finds a bed of coarse reddish tephra and lava deposited between the base of the cones and the large
cinder with temperatures in the range of 38 °C. During the rainy crater walls it is difficult to determine the real altitude of these small
season meteoric waters percolate, get heated and produce a column of structures but they are estimated as follows. The EAC is 80 m high and
steam, which can be easily mistaken as part of the cloudiness often has an ellipsoidal shape with a major axis of 450 m and a minor axis of
surrounding the mountain. 150 m (Fig. 6). At its top it shows two small depressions, remnants of a
small crater. The WEC has a round base of 110 m, a height of 40 m and
4. The deposits from the 1793 eruption of San Martín a small crater on its top (Fig. 6). Thus, the volumes of the former and
latter structures are, approximately, 2.1 × 106 m3 and 0.8 × 106 m3
4.1. Cinder cones and lava flow respectively.
The lava flow extends about 2 km from the rim of the crater and is
From the description of the visit to the crater by Moziño and the composed of lava blocks covered by soil and dense vegetation which
painting by Echeverría, it is clear that the easternmost cone (EAC), obscures its morphology (Fig. 3). The coulee covers approximately an
also the larger of the two, was the active cone. This is also clear from area of 1.4 × 106 m2 filling deep gullies on the northern volcanic

Fig. 8. 1793 lapilli samples seen through a binocular microscope.

 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 197

slopes. Analysis of the change in topography allows estimating its In a series of field seasons during the past years we have observed
thickness in the range between 10 and 20 m yielding an average this layer of lapilli and ash at many locations around the volcano,
volume of the order of 2 × 107 m3. Assuming an average density of except at the far eastern side where it is conspicuously absent. The
2000 kg/m3, the mass is of the order of 4 × 1010 kg. deposit can be correlated by its physical characteristics, its strati-
graphic position, its granulometric characteristics, and its chemical
4.3. Ash fall deposit composition. As stated, in some cases the ash/lapilli is observed on the
surface (e.g. in the S and SW sectors) but more often it is found under
Friedlaender studied the area in the early 1920s and ascended to the modern soil at depths several tens of centimeters. The deposit lies
the crater (Friedlaender and Sonder, 1923). He took a path from San over a brown paleosol and is dark gray to blackish; in some places the
Andrés to Montepio, walked to the east through a pass know as El deposit is bounded by sharp contacts but in others, mostly at the more
Vigía (Fig. 3) and entered the crater from the east, more or less from distant sites, it appears only in lenses or with irregular and faint
the same side that Moziño entered 130 years before. Not far from this contacts (Fig. 7). It contains varying amounts of olivine and pyroxene
point a series of cones trending SE shows long lava flows descending crystals, and glass noticeable to the naked eye or with the hand lens;
to the northeast; based upon the youthful morphologies of the flows, its content of lithics is negligible. The grains are angular to sub-
Friedlaender believed them to have been generated during the 1664 angular, vesicular and frequently fragile. Wohletz (1985) studied the
eruption. Nelson and Gonzalez-Caver (1992) also consider that the mechanism of pyroclast formation in explosive water/magma inter-
long lava flow trending SW and reaching the rim of Laguna Encantada actions and classified pyroclasts on the basis of their morphology,
is from this same eruption. To the NW from El Vigía, not too far from which is related to the different physical mechanisms influencing the
San Andrés there is a cinder cone known as Cerro Vaxin (Fig. 3); in its water/magma interaction. Fig. 8 shows the aspect of lapilli grains
vicinity Friedlaender and Sonder (1923) reported ash-and-lapilli beds under the binocular microscope; these pyroclasts show vesicular,
up to 4 m in thickness, which they attributed to the 1793 eruption. At irregular shapes with fluid form surfaces, and resemble the type II
present, however, the maximum thickness we were able to find in the pyroclasts of Wohletz (1985). Fig. 9 shows Scanning Electron
area is only 1.5 m, although at some places the ash is observed on the Microscope images of pyroclasts from the finer fraction of ash
surface and thus the original deposit might have been thicker. collected at site SM9-2. The grains are very irregular, mostly equant

Fig. 9. SEM images of ash samples from San Martín at site SM9-2. Pyroclasts (a) and (c) resemble type II pyroclasts of Wohletz (1985); (b), (d), (e), and (f) type IV; (g) and (h) type III.
 198
J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208
Fig. 10. Section and sampling points around San Martín volcano and the isopachs (in cm) constructed with measured thickness at locations where the deposit was best preserved.
J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 199

Fig. 11. Histograms of grain-size distributions at some sampling points shown in Fig. 10.
200 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

Fig. 12. Inman coefficients of the deposit at the sampling sites shown in Fig. 10. The data plots within the field of the fall deposits of Walker (1971).

or elongated with moss-like shape and resemble some of the where several layers can be observed; at most locations, however,
pyroclasts type III and IV of Wohletz (1985). such differences are not present and even grading is hardly noticeable.
The paleosol underneath the deposit was sampled at several All samples were sieved for granulometric analysis. Fig. 11 shows
locations and dated with radiocarbon techniques. The conventional the results for several representative samples. Fig. 12 shows a plot of
ages and sampling sites are listed in Table 1 and shown in Fig. 10. The the Inman (1952) coefficients for the deposit on the fields defined by
paleosols have ages ranging between 615 and 290 years BP. Thus, this Walker (1971). The San Martín ash samples with smaller median
layer is beyond doubt of recent origin; which leads to the question diameters show poorer sorting, and all plot within the ash fall field.
about the date in which it was deposited. From the ages found, the We applied the method of Walker (1973) to estimate the style of
deposit could have arisen from the eruptions of 1664 or 1793, since the events that dispersed the ash. To this end the thickness of the
those are the only eruptions thought to have occurred in this period. deposit extrapolated to the origin is 350 cm, thus 0.01 Tmax = 3.5 cm.
From the historical account given before, however, it would seem very Within this isopach the area is of about 450 km2 and the fraction of
unlikely that the deposit had been produced during the 1664 eruption fines (< 1 mm) at the intersection of the 0.1 Tmax (35 cm) with a NW–
with no remaining traces of the 1793 deposit. On the other hand the SE axis is about 60%. Therefore the deposit plots approximately in the
limitations of the C–14 method for dating bulk soils have been amply field of the phreatomagmatic eruptions (Fig. 13).
discussed (Matthews, 1985; McClung de Tapia et al., 2005), and reflect The volume of materials computed from the isopachs by the
the openness of the system and therefore to the lack of knowledge on method of Fierstein and Nathenson (1992), for a single straight line, is
the origin and residence time of organic matter in the soil. Thus, the 2.5 × 108 m3 and –considering an average density of 1.0 g/cm3– a mass
paleosol samples yield ages reliable only within an order of magnitude of ca. 2.5 × 1014 g was calculated. This is probably an underestimation
but attest to the recent age of the deposit. Even when arising from of the actual volume as discussed by Bonadonna·and Houghton
well identified sources the C–14 ages must be taken with care. Sample (2005), who have compared the different methods used to estimate
SM1-16, for instance, is a charcoal sample dated at 290 ± 40 years BP the total volume of tephra ejected in an eruption.
that was collected from a tree log found under the lava flow
descending from the north of San Martín and considered to have
been emplaced during the 1793 eruption; however this age could also 5. Petrography of the eruption products
fit the 1664 eruption. We therefore propose that the ash-and-lapilli
layer discussed above corresponds to the 1793 eruption. In this Petrographic analyses of the products of the eruption were
regard, the 3 cm thick ash layer observed in the core obtained by performed on thin sections from 2 lava samples of the inner cones
Caballero et al. (2006) from Laguna Verde (Fig. 10), which they and 8 ash samples collected at different locations. The results of the
attribute to the 1664 eruption, was more likely deposited in 1793, modal mineral counting are shown in Table B1. In general, the ash
since it is less probable that this eruption had left no traces in the lake shows a larger amount of olivine and a partially altered yellowish
while the minor one in 1664 did.

4.4. Distribution of the ash

Fig. 10 shows the isopachs of the deposit and the results of the
grain-size analyses at selected sites. The isopachs were constructed
using measured thicknesses obtained from the sites where the deposit
was best preserved; in some areas, for instance to the east of the
volcano, the deposit is no longer present because its thickness must
have been very small and it is now embodied in the soil. In Fig. 10 the
sites where ash is observable are indicated by filled circles; however,
only the sites where the deposit seems complete were used for
construction of the isopachs. Notice that the distribution is not
symmetric, a fact not surprising since the main events occurred over
eight months, a period in which the wind velocity and tephra ejection
rate must have varied considerably. These variations are reflected to Fig. 13. Plot of fragmentation vs. dispersal and outline of the fields defined by Walker
some extent in the proximal deposits, for instance at site SM9-9 (1973). The deposits of San Martín plots in the field of phreatomagmatic eruptions.
Table 3
Major and trace element data for 24 selected samples from the 1793 products of San Martín Volcano. The analyses (XRF) were carried out at the Laboratorio Universitario de Geoquímica Isotópica, Instituto de Geología (LUGIS-UNAM).

Sample SM9-6 SM9-14 SM1-4 SM1-5 SM2-1 SM2-5 SM2-25 SM2-30 SM2-37 SM2-45a SM2-45b SM3-18 SM3-19 SM3-21 SM3-22a SM3-22b- SM4-7 SM4-16a SM4-16b SM4-16c SM9-2 SM6-3 SM6-4

J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208
Rock type Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Ash Lava Lava

Latitude N 18.545 18.639 18.637 18.549 18.631 18.596 18.569 18.598 18.551 18.521 18.521 18.516 18.517 18.527 18.507 18.507 18.594 18.528 18.528 18.528 18.549 18.563 18.562

Longitude W 95.240 95.246 95.200 95.231 95.233 95.234 95.281 95.300 95.278 95.160 95.160 95.204 95.203 95.207 95.2008 95.2008 95.200 95.209 95.209 95.209 95.231 95.201 95.199

Mayor oxides (wt%)

SiO2 45.47 45.26 45.26 45.96 45.17 44.59 44.48 46.24 44.57 42.37 43.52 45.68 44.22 46.07 45.12 44.89 44.70 45.01 45.36 44.70 45.04 48.15 48.25
TiO2 1.62 1.41 1.41 1.50 1.37 1.51 1.50 1.33 1.58 1.83 1.82 1.59 1.51 1.34 1.52 1.50 1.61 1.78 1.59 1.63 1.51 1.49 1.47
Al2O3 14.52 14.05 13.65 13.72 13.65 14.32 13.91 13.39 13.48 11.97 14.15 13.74 13.93 14.26 13.78 13.82 13.91 13.23 13.29 14.15 13.12 15.53 15.28
Fe2O3(t) 11.83 11.46 11.81 11.48 11.58 11.84 11.81 11.32 12.14 12.87 12.50 11.87 11.72 11.18 11.61 11.83 12.16 12.40 12.00 12.08 11.45 10.43 10.49
MnO 0.18 0.17 0.18 0.17 0.17 0.18 0.18 0.17 0.18 0.18 0.18 0.18 0.19 0.18 0.17 0.18 0.17 1.18 0.17 0.18 0.16 0.16 0.16
MgO 12.63 14.08 13.99 13.77 14.96 14.66 14.61 13.88 14.96 17.93 13.87 13.18 14.93 12.34 13.92 13.94 14.20 13.84 14.00 13.80 14.24 9.01 9.54
CaO 9.75 9.74 9.32 9.85 9.59 9.26 9.56 10.32 9.46 8.89 9.73 9.64 9.53 10.37 10.12 9.71 9.51 9.26 9.44 9.72 9.87 10.03 10.19
Na2O 2.45 2.07 2.15 2.40 2.06 2.08 2.11 2.04 1.98 2.15 2.54 2.63 2.24 2.39 2.47 2.34 2.30 2.56 2.46 2.50 2.30 3.18 3.05
K2O 1.03 0.84 0.87 0.98 0.83 0.84 0.87 0.78 0.82 0.99 1.07 1.03 0.92 0.86 0.98 0.94 0.91 1.03 0.97 1.02 .95 1.16 1.08
P2O5 0.39 0.41 0.34 0.35 0.39 0.36 0.36 0.39 0.38 0.50 0.51 0.42 0.39 0.44 0.37 0.41 0.42 0.49 0.43 0.45 0.33 0.49 0.46
L.O.I. 0.06 0.20 0.29 − 0.14 − 0.08 0.38 0.20 0.07 0.45 − 0.03 − 0.04 − 0.12 − 0.27 0.20 0.24 − 0.1 0.35 − 0.29 0.18 − 0.28 0.30 − 0.23 − 0.23
Total 99.92 99.70 99.49 100.12 99.69 100.01 99.60 99.93 99.99 99.65 99.96 99.85 99.31 99.63 100.31 99.46 100.14 99.51 99.91 99.95 99.47 99.40 99.74

Trace elements (ppm)

Ni 360 464 480 469 516 530 591 498 601 1200 476 434 631 352 501 507 531 555 537 498 424 154 154
V 315 264 252 286 240 276 229 238 228 232 260 253 203 269 272 237 268 182 214 274 354 261 278
Cr 862 908 839 1050 960 1032 1049 817 1000 1097 768 729 951 694 956 839 932 751 837 851 975 440 435
Co 63 77 77 70 77 78 87 78 88 161 77 73 94 60 80 82 85 91 82 82 57 41 41
Cu 98 97 81 105 105 97 111 83 97 105 100 93 113 77 116 100 100 93 32 106 95 72 61
Zn 111 115 115 111 114 115 123 120 127 198 130 122 132 109 118 126 128 147 130 131 103 90 94
Rb 18 18 21 19 16 17 19 18 18 29 22 24 22 17 20 22 20 25 23 22 19 23 23
Sr 648 679 636 613 663 596 611 752 619 738 729 698 656 819 636 672 657 694 665 720 609 864 781
Y 24 26 25 26 27 27 29 30 29 46 27 25 30 26 26 29 29 32 28 21 19 21 21
Zr 170 163 167 158 161 164 171 162 173 248 193 179 177 173 160 175 176 201 184 107 139 171 160
Nb 23 16 17 19 16 21 19 13 19 28 23 20 18 15 17 19 21 24 22 21 17 15 13
Ba 24 285 281 265 304 251 242 328 266 327 314 342 273 359 274 301 254 318 311 307 244 688 504
Pb 5 4 5 2 5 3 3 6 3 4 2 7 2 6 3 3 5 6 3 6 5 6 2
Th 7 3 4 5 5 4 1 6 4 4 5 7 4 2 2 4 3 7 7 6 3 8 7

201
202 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

matrix glass, while the cones show larger amounts of plagioclase

phenocryts and microphenocrysts. Detailed descriptions of the thin
sections are given in Appendix B.

6. Chemical and isotopic composition

From the collected ash samples we selected 21 from the best

preserved deposits around the volcano and two lava samples, each
one from the EAC and WEC cones, for major and trace element
analyses via X-ray fluorescence. The results are shown in Table 3. Sr,
Nd, and Pb isotopic analyses were carried out on 10 samples (Table 4).
Analytical procedures for XRF and isotopic data are given in Lozano-
Santa Cruz et al. (1995) and Schaaf et al. (2005), respectively. In the
following discussion data of Nelson et al. (1995) and Verma (2006) for
samples belonging to the YVS are included for comparison.
In the Total Alkalies vs. Silica (TAS) diagram (Fig. 14), most of the
samples plot near the limits between basic and ultrabasic rocks with
SiO2 contents ranging between 46.2 and 42.4 wt.% (ash) and 48.2 wt.%
(lavas). In general, the differences between ash and lava samples are
due to the higher weight percent of Fe2O3(t), MnO, and MgO in ash
and the opposite behavior for the rest of the oxides (Table 3). It is
noticeable that both lavas have practically the same chemical
Fig. 14. San Martín 1793 products in a plot of total alkalis versus silica the data of Nelson
composition and that the average of the ash leans towards the et al. (1995) have been included for comparison.
ultrabasic field. This is probably due to the atmospheric selection of
the grains during deposition since the less dense glass shards can be
transported to longer distances in the atmosphere. Almost all of the accompanied by minor assimilation of crustal material and/or
samples plot in the alkaline field of the TAS diagram and in the same sediments. Nelson's data show nearly identical 87Sr/86Sr but less
area as the rocks analyzed by Nelson et al. (1995). The samples of evolved εNd in comparison to our samples. This may be due again to
these authors show, however, a greater dispersion probably because the different sample types, ours being more recent ash and lava. 87Sr/
86
they belong to rock types of different origins in the YVS. Sr ratios are less radiogenic than most Mexican Volcanic Belt
Fig. 15 shows Harker diagrams of oxides vs. MgO. The cones and samples such as Popocatépetl (Schaaf et al., 2005) at nearly identical
ash plot in about the same area as the samples of Nelson et al. (1995), εNd values. This can be best explained by variable assimilation ratios
indicating that no significant change in chemistry has occurred during of different crustal components. In the case of Popocatépetl volcano,
the history of the YVS. Note also that the samples from the cones shallow carbonate material is involved; at San Martín volcano a
appear separated from the ash but there is no observable systematic lithologically different crusts seems to form the contaminating
difference between both cones. MgO contents are between 12 and component.
18 wt.% for ash samples. Decreasing CaO with increasing MgO When plotting 207Pb/204Pb vs. 206Pb/204Pb (Fig. 17) San Martín
indicates increasing olivine fractionation whereas decreasing Al2O3, samples are placed between typical values for the East Pacific Rise
Na2O, and Ba yields evidence for decreasing plagioclase fractionation; (EPR) MORB, and ocean floor sediments from the Pacific Ocean and
yet it is difficult to relate this to a real difference between the first and material from the Paleozoic Acatlán Complex of Central Mexico. This
the last products of the eruption or to the processes of separation in evidence suggests crustal assimilation from these sources and rules
the ash cloud. out possible contributions from the Grenvillian Oaxaca Complex with
Strontium (Sr) and Neodymium (Nd) isotopic data displayed in considerably less radiogenic Pb isotopes. The considerable spread in
207
Fig. 16 show the San Martín samples in the mantle array field Pb/204Pb at nearly identical 206Pb/204Pb gives evidence for the

Table 4
Isotopic analyses for selected samples from the 1793 eruption of San Martín volcano Sr–Nd isotopes, lead isotopes ratios.
87
Sample Rock type Sr/86Sr 1sd* 143
Nd/144Nd 1sd* εNd 1sigma 206
Pb/204Pb (%std err) 207
Pb/204Pb (%std err) 208
Pb/204Pb (%std err)

SM9-2 Ash 0.703228 30 0.512950 17 6.09 0.33 na na na na na Na

SM9-14 Ash 0.703334 38 0.512930 19 5.70 0.37 18.7379 0.071 15.5724 0.078 38.4301 0.087
SM2-1 Ash 0.703341 38 0.512931 20 5.72 0.39 18.7459 0.053 15.5785 0.059 38.4493 0.065
SM2-10 Ash 0.703425 39 0.512910 19 5.31 0.37 na na na na na na
SM2-25 Ash 0.703232 38 0.512954 15 6.16 0.29 18.7519 0.053 15.5670 0064 38.4114 0.079
SM3-10 Lava 0.703314 39 0.512916 13 5.42 0.25 18.7374 0.020 15.5643 0.078 38.4234 0.021
SM3-11 Lava 0.703327 47 0.512954 18 6.16 0.35 18.7397 0.045 15.6254 0.063 38.4372 0.054
SM3-12 Lava 0.703255 34 0.512949 19 6.07 0.37 18.8880 0.090 15.6417 0.103 38.6025 0.092
SM6-3 Lava 0.703424 31 0.512940 18 5.89 0.35 18.7470 0.018 15.5919 0.020 38.4960 0.022
SM6-4 Lava 0.703430 32 0.512938 19 5.85 0.37 18.7458 0.015 15.5922 0.016 38.4949 0.016

na — Not analyzed.
Sr, Nd and Pb isotope ratios were determined using a Finnigan MAT262 mass spectrometer equipped with 8 variable Faraday collectors, installed at the Laboratorio Universitario de
Geoquímica Isotópica (LUGIS) at the Universidad Nacional Autónoma de México (UNAM). The isotopic measurements were made in static collection mode. Sr and Nd were loaded as
chlorides on rhenium filaments. 2σ(m) values (2σ(m) or 2 SE(M) = 2σabs √n) were calculated from approx. 60 individual isotopic determinations for Sr and Nd. The measured 87Sr/
86
Sr values were normalized to a 86Sr/88Sr value of 0.1194, and those of 143Nd/144Nd to a 146Nd/144Nd ratio of 0.7219. 87Sr/86Sr of the NBS987 and 143Nd/144Nd of the La Jolla-standard
throughout this study were 0.710235 ± 18 (1σabs, n = 237) and 0.511878 ± 21 (1σabs, n = 134). Pb samples were loaded with a mixture of silica gel and phosphoric acid and runs
consisted of approx. 100 individual measurements. Pb ratios were corrected for 0.13% fractionation per mass unit by comparing with the mean value of 23 measurements of NIST-
NBS981 (Pb) (206Pb/204Pb = 16.895 ± 0.05%, 207Pb/204Pb = 15.4314 ± 0.08%, 208Pb/204Pb = 36.5217 ± 0.014%). ⁎Standard deviation.
Total procedure blanks were 0.57–31 ng for Sr and 1.05–22 ng for Nd. Chemical blank for Pb was 94 pg.
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 203

Fig. 15. Harker diagrams for major and trace elements against MgO for the 1793 San Martín products. Data of Nelson et al. (1995) are also plotted for comparison.
204 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

that in the copy of Moziño's report conserved in the AGI the presence
of a spring inside the crater is mentioned; therefore abundant
groundwater was available to take place in a possible phreatomag-
matic interaction. On the other hand, evidence of phreatomagmatic
activity such as the presence of accretionary lapilli and base surge
beds, which are common in andesitic phreatomagmatic eruptions,
could not be identified in the deposits. However, accretionary lapilli,
due to its weight, could have been deposited closer to the vent,
probably on the slopes of the volcano, which is now largely covered by
vegetation. Moreover, the vent, which eventually formed the EAC and
perhaps the WEC, was shielded by the large crater of San Martín,
whereby the main surge deposits were probably deposited inside this
crater. Echeverria's painting shows dunes around the erupting vent
which evoke this type of deposit.
A more recent example of the morphology of pyroclasts derived
from phreatomagmatic explosions was presented by Taddeucci et al.
(2004) for products of the July–August 2001 explosive activity of Mt.
Etna. Some of the pyroclasts produced during this period of activity
are nonvesicular, blocky, and show hydration cracks that were
produced by hydromagmatic activity; some others show well
Fig. 16. 87Sr/86Sr vs. εNd plot for rocks of the TVF compared with volcanic rocks from developed, spherical, and occasionally coalescing vesicles that were
Popocatépetl, Mexico (Schaaf et al., 2005), and MORB. Data of Nelson et al. (1995) and
Verma (2006) are also plotted for comparison.
produced during phases of Strombolian activity. The pyroclasts
produced in 1793 by San Martín are likewise of these two types,
since part of the ashes show vesicularity, are elongated and delicate
presence of young ocean floor sediments in the TVF magmas (Schaaf suggesting their generation during a later phase of Strombolian
and Carrasco-Núñez, 2010-this issue). activity, which is depicted in Echeverría's painting and described by
Moziño.
7. Discussion The height of the plumes produced during the explosions must
have been several kilometers, since according to Moziño the ash
The beginning of the eruptive period at San Martín volcano, a rapid clouds of the 22 March eruption reached Oaxaca (located 250 km SW
succession of strong explosions heard at distances of 300 km, is from the volcano) the following day. How long the ash falls continued
strongly reminiscent of hydrovolcanic explosions as described by is not reported by the naturalist, but he related that being in Oaxaca at
Wohletz (1985). Explosions of this kind are produced by flash the time he witnessed the episode and was able to comfort his
vaporization of water due to rapid energy transfer from the magma acquaintances worried by such an event. In its way to Oaxaca the ash
by superheating. The mechanical energy derived from the vaporiza- cloud passed over the topographic barrier presented by the Sierra
tion is then partitioned into several dominant modes including: Madre Oriental with a maximum elevation of about 2600 m (Fig. 18).
fragmentation energy, particle kinetic energy, seismic energy, and This suggests that the eruptive column reached an altitude of several
acoustic energies. The eruptive phases occurring on March, May and kilometers. To estimate its possible height we applied the diffusive
April 1793 were probably of this type. Several facts support this model of Suzuki (1983) considering that at Oaxaca the thickness of
hypothesis: first, they began suddenly, without any warning activity the deposit must have been at least of a few millimeters since it
(e.g. felt seismicity). Second, they produced abundant millimetric provoked concern among the people. As input for the model we
pyroclasts and were distributed widely (Fig. 13). Third, the morphol- considered a plume with a mass of 5 × 1013 g (about 1/5 of the total
ogy of a significant part of the pyroclasts is irregular and poorly mass estimated), a height of 12 km, and a lognormal particle
vesiculated (Fig. 9). Fourth, as was mentioned before, the edifice of distribution with median diameter 0.47 and standard deviation of
San Martín possesses many springs with abundant yield that satisfies 0.4; values that fit roughly the total granulometry of the deposit. In
the needs of the towns around the volcano. It is interesting to note addition, we considered the wind speeds measured at nearby
Veracruz by the National Meteorological Service, which are as large
as 10 m/s at altitudes of 6 km and 37 m/sec at 11 km asl. The model
with the specified data yields a thickness of a few millimeters at a
distance of 250 km on the dispersion axis. Although we do not know
either the dispersion axis of the deposit left by that single eruption, or
some other parameters of the eruption, the values chosen are realistic
and suggest that the height of the column was probably in the range of
10 km due to the height of the volcano (1659 masl).
The final part of the eruptive period, which probably lasted two
more years, was effusive and formed a lava flow. Eruptions of this type
must have occurred before, since aerial photographs show several
flows descending from the flanks of the volcano.
The petrologic and geochemical analyses of the 1793 products are
not focused on solving the question of the tectonic origin of the TVF,
but to provide a more complete characterization of these products.
Our results show that the most recent products of the volcanic activity
Fig. 17. 207Pb/204Pb vs. 206Pb/204Pb plot for volcanic products of the TVF as compared to in the TVF are not significantly different from other rocks of the YVS.
other rocks from Mexico: (NHRL) North hemisphere reference line. (AC) Acatlan
complex. (OC) Oaxaca complex. (POS) Pacific Ocean sediments. (EPR) East Pacific Rise
The ascent of this primitive, mantle derived magma to the surface is
(Modified after Martínez-Serrano et al, 2004). SK is the lead evolution line of Stacey and aided by a tensional stress field, attested by a strong positive gravity
Kramers (1975). anomaly associated to the field (De la Fuente et al., 1994). A positive
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 205

Fig. 18. Topographic cross section across Oaxaca and San Martín volcano (AA', Fig. 4). The ash from the 1793 eruption reached the city of Oaxaca and therefore in its descent had to
surpass the height of the western mountainous range.

gravity anomaly indicates that dense material intruded units of less Hernández, Tomás, José and Joaquin Palacios, Angel Cota, and Pablo
dense rocks; this requires a tensional field facilitating the passage of Solis from Mpo. San Andrés Tuxtla and Don Eneas Rivas Castellanos,
such material. The gravity anomaly has a long axis trending SE, the chronicler of Santiago Tuxtla, are thanked for their help during our
same direction of the lineaments formed by the location of the field work. Professor A. Udias and Dr. Asmaa Bouhrass, kindly guided
parasitic cones of San Martín volcano and other monogenetic cones in AZC to Moziño's documents in the “Archivo General de Indias”.
the YVS of the TVF. Fruitful discussions with Drs. Raimundo Martínez and José Luis Bravo
Houghton et al. (2004) have made important considerations on are kindly acknowledged. Finally, we are very thankful to Claus Siebe,
basaltic plinian eruptions that are applicable to other less explosive an anonymous reviewer, and a guest editor Lee Siebert for their
basaltic eruptions as well. These authors observe that plinian basaltic careful review of the manuscript; their comments helped us to correct
eruptions are poorly understood due to their relative scarcity in the several mistakes and improve the manuscript.
geologic record. They also point out that, due to the fast ascent rates of
basaltic magma, the warning times are shorter than for other type of Appendix A. Sources of the report of Don Joseph Mariano Moziño
eruptions. This is of great importance in the case of San Martín volcano on the 1793 eruption of San Martín Volcano
and the whole TVF since the crust in the region is only 26 km thick
(Zamora-Camacho, 2007). Moreover, since these eruptions are less The text of Don José Mariano Moziño report consulted for this
frequent than other types of basaltic eruptions the warning signals work was obtained from a copy sent to the King of Spain by the
can be misunderstood or ignored. In this regard, the 1793 eruption of Viceroy of New Spain, Don Juan Vicente de Güemes Pacheco de Padilla
San Martín is a good example since it began without warning signals y Horcasitas, the second Count of Revillagigedo. The manuscript of this
that could be perceived by the nearby population, who even mistook copy is conserved at the AGI and is titled “Descripción del volcán de
the first explosions for a thunderstorm. Tuxtla. Por D. José Mariano Moziño. Botánico de la expedición en N.E.
Año de 1793”. We also compared the version published by the SMGE.
8. Conclusions The text of both versions is similar, except for a few paragraphs and
numbers. For example the SMGE version states that the thickness of
Moziño's description of the 1793 eruption of San Martín volcano the ash bed around San Martín was 3 “varas”, whereas the manuscript
and our field work suggest that the events of March, May and June at the AGI reports 2 “varas”. The later version mentions as well the
consisted of phreatomagmatic explosions, which were followed by existence of a spring inside the crater of San Martín that is not
Strombolian phases that ended in October 1793. The eruption mentioned in the former.
continued with a mostly effusive stage that lasted probably 2 years. A translation to English of the report can be obtained from:
The explosive phases produced ash falls of more than 1 cm in http://www.geofisica.unam.mx/vulcanologia/spanish/personal/
thickness over an area of at least 480 km2. During the activity at least espindola.html.
2 × 1011 kg of ash and 2 × 1010 kg of lava were erupted. A new period
of activity at San Martín volcano is probable. Furthermore, considering Appendix B. Petrography of the products from the inner cones,
that the products emitted in the TVF are primitive basaltic magmas of and ash
low viscosity and the relatively thin crustal thickness in the area, such
activity would happen with little warning time. B1. Lava cones

Acknowledgments B1.1. East Cone (EAC)

Porphyritic texture with phenocrysts of plagioclase (Plg), olivine
This work was carried out with support from UNAM grant PAPIIT (Ol) and clinopyroxene (Cpx) in a cryptocrystalline matrix in which
IN105003-3 and CONACYT grant 428427-F. The authors are indebted microlites, opaque minerals (oxides) and glass can be recognized.
to Gabriela Solís, Teodoro Hernández, Juan Morales, Rubin Lozano and Plagioclase phenocrysts (0.3–1.6 mm long) are tabular, euhedral, and
Patricia Girón from LUGIS-UNAM for isotopic and geochemical show Carlsbad twinning. Zoning is scarce and when present is not
analyses. Margarita Reyes and Blanca Sonia Angeles from Instituto conspicuous. Olivine crysts show internal corrosion marks and glass
de Geología were very helpful with the SEM and Diego Aparicio, inclusions are partially or totally replaced by opaque minerals
Enrique Gonzalez, and Renato Castro with the preparations of thin (Fig. B1a, c). Dissolution embayments are rare but some glomer-
sections and petrographic analyses. Sres. Celso, Álvaro and Lamberto ocrysts are present. In a few Plg crysts it is possible to observe apatite
206 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

Fig. B1. Thin sections under polarizing light of EAC sample SM6-4. Notice in: (a) the zoning of Cpx, (b) the growth of Plg, (c) the corrosion of Ol, (d) the zoning in Plg.

(< 0.05 mm long). The microcrysts of Plg (0.03–0.3 mm) are tabular, B.1.2. West cone (WEC)
euhedral, almost without zoning and rarely show signs of corrosion. This thin section is mostly very similar to the former (EAC) except
Ol-phenocrysts (0.3–1.2 mm long) are subhedral and do not show that it has a larger content of Cpx phenocrysts and higher crystallinity.
zoning. Chrome spinel inclusions (< 0.07 mm) are common in the The Ol/Cpx ratio is < 1 while in the EAC it is greater than 2 (Table B1).
olivines while Plg inclusions (< 0.2 mm) are rare. The Ol micro- The rock is porphyritic with phenocrysts of Plg, Ol and Cpx in a
phenocrysts (0.03–0.3 mm long) are similar to the phenocrysts but cryptocrystalline matrix, containing Plg microlites, and opaques. Plg
show less corrosion and no Plg inclusions. Phenocrysts (0.3–2.0 mm) phenocrysts (0.3–2.0 mm) are tabular, euhedral, and show Carlsbad
of Cpx (augite) are greenish to light brown, with prismatic habits, and twinning. Slight zoning is common and more evident than in the EAC
show some internal corrosion. Simple twinning is common and section. Internal corrosion is frequent and more extended, mainly in
different types of moderate zoning are observable: oscillating, the largest crystals, in comparison to the EAC. Glass inclusions are
sectorial, and hourglass (Fig. B1b,d). Glomerocrysts of Cpx associated partially or completely replaced by opaque minerals. Dissolution
with Ol are frequent. Cpx microphenocrysts (0.03–0.3 mm) are rarely embayments are also rare and some glomerocrysts can be observed.
zoned and show only slight signs of corrosion. Microphenocrysts of Plg (0.03–0.3 mm) are tabular, euhedral, almost

Table B1
Modal mineralogical analyses of 1793 San Martín lava and ash samples (in vol.% different phases): phenocrystals (p) > 0.3 mm; microphenocrystals (mp) 0.03–0.3 mm) and vesícles
(Ves.). For the WEC and EAC lavas 1150 points were counted and 1000 for the ash. Plg, Ol, Cpx, amphibole and opaque minerals were considered.

Sample SM6-4 (WEC) SM6-3 (EAC) SM1-5 SM3-18 SM3-21 SM2-37a SM2-30 SM2-1 SM9-14a SM2-25

Rock type Lava Lava Ash Ash Ash Ash Ash Ash Ash Ash

Plg (P) 13.18 14.04 1.87 0.81 5.21 0.31 1.60 0.50 0 0.60
Plg (mp.) 19.33 22.92 3.66 3.34 5.51 6.22 5.32 1.91 4.18 3.17
Ol (P) 8.715 9.36 9.96 14.59 15.13 7.55 18.2 14.30 11.60 4.71
Ol (mp) 2.46 3.51 7.32 4.66 3.01 5.61 4.31 2.52 3.24 5.12
Cpx (P) 4.02 10.06 0.81 3.55 9.52 0.31 10.7 0 0 1.33
Cpx (mp) 4.25 4.68 1.02 4.15 5.21 1.33 7.32 0.20 0 1.02
Amphíbole 0 0 0 0 0.30 0 0.10 0 0 0
Opaque 1.23 1.29 0.41 0.20 0.70 1.02 1 0.40 0.73 0.92
Matrix Alt.* 0 0 48.68 46.10 41.88 34.90 44.60 57.40 43.10 54.20
Matrix P.A.** 46.70*** 34.15*** 26.52 22.59 13.53 42.80 6.82 22.70 37.20 28.90
TOTAL 99.885 100.01 100.25 99.99 100 100.05 99.97 99.93 100.05 99.97

*The matrix is altered to opaque mineral. **PA partially altered matrix (altered to a brown-yellowish minerals in the case of the ash). ***The matrix of WEC and EAC lavas is
cryptocrystalline with Plg microlites and opaque minerals.
 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208 207

Fig. B2. (a) to (c) Thin section of WEC sample SM6-3 (CO). Notice the following features: (a) zoning of Plg and corrosion of Cpx and Ol, (b) and c) strong corrosion of Ol with
inclusions of Sp, (d) Crystals of Ol and Plg in ash sample SM9-2.

without zoning, and rarely exhibit signs of corrosion (Fig. B1a). Ol euhedral, and commonly shows spinel inclusions (Figs. B1 and B2). Ol
phenocrysts (0.3–1.5 mm) are subhedral, show no zoning, and crystals show no zoning and the internal corrosion varies from weak
moderate to high internal corrosion. As in the EAC sample only Plg to severe. Cpx crystals are light olive green displaying weak
shows rare tabular apatite inclusions. Chrome spinel inclusions pleochroism, generally subhedral and showing twinning, particularly
(< 0.09 mm) are common (Fig. B1b). Ol-microphenocrysts (0.03– simple twinning. Moderate internal corrosion is present although not
0.3 mm) of are similar to those in the EAC but show lower degrees of to the extent of developing a mesh texture. Glomerocrysts of Cpx are
corrosion. Cpx phenocrysts (augite, 0.3–3.0 mm) are greenish to light observed in some ash grains.
brown, prismatic and frequently with signs of internal corrosion. Cpx The microphenocrysts of Plg, Ol and Cpx do not show the same
crystals also show moderate sector or hourglass zoning, and frequent degree of corrosion of the phenocrysts, being almost nil or very slight
simple twinning. Opaque mineral inclusions are less common than in in most cases. Plg and Cpx microcrysts show generally no zoning.
Ol (Fig. B1c). Glomerocrysts of sole Cpx or in combination with Ol are Only in two cases (SM3-21, SM2-30, Table B1) were amphiboles
frequent. Cpx microcrysts rarely exhibit zoning and signs of corrosion. observed. They are brown, anhedral, show corrosion on their margins,
which makes them subrounded, and reaction rims with another
B.2. Ash samples crystalline phase. Their reduced number suggests an exogenous origin.

Ash samples show the same mineral association as the cone

samples; however their Ol content is always larger than the Cpx References
crystal content. The proportions of Plg, Ol, and Cpx crystals vary in the
Aguilera Gómez, L., 1988. Petrología de las rocas ígneas del área de los Tuxtlas, Veracruz.
different samples; for example samples SM2-1 and SM9-14 contain Tesis Profesional, Instituto Politécnico Nacional, Mexico, 58 pp.
almost no Cpx (Table B1). Shattered crystals are common; however, Bonadonna, C., Houghton, B.F., 2005. Total grain-size distribution and volume of tephra-
under the microscope one can recognize a porphyritic texture with a fall deposits. Bull. Volcanol. 67, 441–456.
Caballero, M., Vázquez, G., Lozano-García, S., Rodríguez, A., Sosa-Nájera, S., Ruiz-
matrix of varying degrees of alteration. The matrix is dark to light Fernández, A.C., Ortega, B., 2006. Present limnological conditions and recent (ca.
brown and greenish brown to completely altered with abundant 340 yr) palaeolimnology of a tropical lake in the Sierra de Los Tuxtlas, eastern
opaque minerals. In the partially altered matrix, tabular Plg crysts Mexico. J. Paleolimnol. 35, 83–97.
De la Fuente, M., Mena, M., Aiken, C., 1994. Cartas gravimétricas de la Republica
(< 0.1 mm long), glass, opaque minerals and some tabular Cpx crysts Mexicana. Instituto de Geofísica, UNAM.
(< 0.1 mm long) can be observed (Fig. B2d). Some of the tabular Engstrand, I.H.W., 1981. Spanish Scientists in the New World: The Eighteenth-Century
crystals show a preferred orientation. Plg-phenocrysts and micro- Expeditions. Seattle and London: Washington University Press, N.Y.
Felix, J., Lenk, H., 1890. Beitraege zur Geologie und Palaeontologie der Republik Mexico.
phenocrysts of are tabular, euhedral, and not abundant. Internal
Verlag Arthur Felix, Leipzig. 142 p.
corrosion in part of the crystals is relatively frequent although rarely Fierstein, J., Nathenson, M., 1992. Another look at the calculation of fallout tephra
forming a mesh texture. The Ol, when not fragmented, is subhedral or volumes. Bull. Volcanol. 54, 156–167.
208 J.M. Espíndola et al. / Journal of Volcanology and Geothermal Research 197 (2010) 188–208

Friedlaender, I., Sonder, R.A., 1923. Über das Vulkangebiet von San Martín Tuxtla in Pieschel, C., 1855. Die Vulkane der Republik Mexiko, Zeitschrift für allgemeine
Mexiko. Zeit. Vulkan. 7, 162–187. Erdkunde, B. 4, 379–399.
García, J. A., 1835. Eruptionen des Vulkanes on Tustla in den Jahren 1664 und 1793. El Pichler, H., Weyl, R., 1976. Quaternary alkaline volcanic rocks in eastern Mexico and
Constitucional (Jalapa 8 Dic. 1830). Neues Jahrbuch für Mineralogie, 6, 40–456. Central America. Munster. Forsch. Geol. Paläont. 38/39, 159–178.
Houghton, B.F., Wilson, C.J.N., Del Carlo, P., Coltelli, M., Sable, J.E., Carey, R., 2004. The Reinhardt, B.K., 1991. Volcanology of the younger volcanic sequence and volcanic
influence of conduit processes on changes in style of basaltic Plinian eruptions: hazards study of the Tuxtla volcanic field, Veracruz, Mexico, M.Sc. Thesis, Tulane
Tarawera 1886 and Etna 122 BC. J. Volcanol. Geotherm. Res. 137, 1–14. University, 147 pp.
Inman, D.L., 1952. Measures for describing the size distribution of sediments. J. Sediment. Sapper, K., 1917. Katalog der geschichtlichen Vulkanausbrüche. Karl J. Trübner,
Petrol. 22, 125–145. Straβburg. p. 244.
Lamb, H.H., 1970. Volcanic dust in the atmosphere; with a chronology and assessment Schaaf, P., Carrasco-Nuñez, G., 2010. Geochemical and isotopic profile of Pico de Orizaba
of its meteorological significance. Philos. Trans. R. Soc. A 266, 425–533. (Citaltepetl) volcano, Mexico: insights for magma generation processes. J. Volcanol.
Lozano-Santa Cruz, R., Verma, S., Girón, P., Velasco, F., Morán-Zenteno, D., Viera, F., Gotherm. Res. 197, 108–122.
Chávez, G., 1995. Calibración preliminar de fluorescencia de rayos X para análisis Schaaf, P., Stimac, J., Siebe, C., Macías, J.L., 2005. Geochemical evidence for mantle origin
cuantitativo de elementos mayores en rocas ígneas. Instituto Nacional de and crustal processes in volcanic rocks from Popocatépetl and surrounding
Geoquímica, México, Actas 1, 203–208. monogenetic volcanoes, central Mexico. J. Petrol. 46, 1243–1282.
Matthews, J., 1985. Radiocarbon dating of surface and buried soils: principles, problems Simkin, T., Siebert, L., 1994. Volcanoes of the World. Smithsonian Institution,
and prospects. In: Richards, K.S., Arnett, R.R., Ellis, S. (Eds.), Geomorphology and Geoscience Press Inc., Tucson, Arizona. 349 pp.
Soils. Allen and Unwin, London, pp. 271–288. Stacey, J.S., Kramers, J.D., 1975. Approximation of terrestrial lead isotope evolution by a
Martínez-Serrano, R.G., Schaaf, P., Solis-Pichardo, G., Hernández-Bernal, M.S., Hernández- two-stage model. Earth Planet Sci. Lett. 26, 207–221.
Treviño, T., Morales-Contreras, J.J., Macias, J.L., 2004. Sr, Nd and Pb isotopes and Suzuki, T., 1983. A theoretical model for the dispersion of tephra. In: Shimozuru, D.,
geochemical data from the Quaternary Nevado de Toluca volcano and the Tenango Yokoyama, I. (Eds.), Arc Volcanism: Physics and Tectonics. Terrapub, Tokyo, pp. 95–113.
Volcanic Field, Mexico, a source of recent adakitic magmatism. J. Volcanol. Geotherm. Taddeucci, J., Pompilio, M., Scarlato, P., 2004. Conduit processes during the July–August
Res. 138, 77–110. 2001 explosive activity of Mt. Etna (Italy): inferences from glass chemistry and
McClung de Tapia, E., Dominguez Rubio, I., Gama Castro, J., Solleiro, E., Sedov, S., 2005. crystal size distribution of ash particles. J. Volcanol. Geotherm. Res. 137, 33–54.
Radiocarbon dates from soil profiles in the Teotihuacan Valley, Mexico: indicators Thorpe, R.S., 1977. Tectonic significance of alkaline volcanism in eastern Mexico.
of geomorphological processes. Radiocarbon 47, 159–175. Tectonophysics 40, T19–T26.
Medel y Alvarado, L., 1963. Historia de San Andrés Tuxtla, Veracruz, 1532–1950, Verma, S.P., 2006. Extension related origin of magmas from a garnet-bearing source in
Primera edición facsimilar, 1993. Gobierno del Estado de Veracruz, México. vol. 1, Los Tuxtlas volcanic field, Mexico. Int. J. Earth Sci. 95, 871–901.
pp. 13–132. Walker, G.P.L., 1971. Grainsize characteristics of pyroclastic deposits. J. Geol. 79,
Mooser, F., Meyer-Abich, H., McBirney, A.R., 1958. Catalogue of the active volcanoes and 696–714.
solfatara fields of Central America. In: International Volcanological Association Waitz, P., 1915. Der gegenwärtige stand der Mexikanischen Vulkane. Zeit. fur Vulkan.
(Ed.), Catalogue of the Active Volcanoes of the World, Naples, pp. 32–34. Band I, 259–268.
Moziño, J.M., 1870. Informe sobre la erupción del Volcán de San Martín Tuxtla Walker, G.P.L., 1973. Explosive volcanic eruptions — a new classification scheme. Geol.
(Veracruz) ocurrida el año de 1793. Bol. Soc. Mex. Geogr. Estad. 2, 62–72. Rundsch. 62, 431–446.
Mühlenpfordt, E., 1844. Versuch einer getreuen Shilderung der Republik Mejico, Wohletz, K.H., 1985. Mechnisms of hydrovolcanic pyroclast formation: grain-size,
besonders in Beziehung auf Geographie, Ethnographie und Statistik. Bd. 1–2, scanning electron microscopy, and experimental studies. J. Volcanol. Geotherm.
Hannover. Res. 17, 31–63.
Nelson, S.A., Gonzalez Caver, E., 1992. Geology and K–Ar dating of the Tuxtla Volcanic Zamora-Camacho, A., 2007. Estudio vulcanológico del área del volcán San Martín
Field, Veracruz, Mexico. Bull. Volcanol. 55, 85–96. Tuxtla, Veracruz, México, Ph. D. Thesis, UNAM, México.
Nelson, S.A., Gonzalez Caver, E., Kyser, T.K., 1995. Constraints on the origin of alkaline
and calc-alkaline magmas from the Tuxtla Volcanic Field, Veracruz, Mexico.
Contrib. Mineral. Petrol. 122, 191–211.