B9CT

FI

lv.2:2|
cop. 3

UN* RSITY OF
ILLIi^iS LIBRARY

ATURBANA-CHAMPAIGN
NATURAL HIST. SURVEY

FIELDIANA • GEOLOGY

Published by
CHICAGO NATURAL HISTORY MUSEUM

Volume 10 August 30, 1957 No. 30

A RESTUDY OF THE 1917 ERUPTION
OF VOLCAN BOQUERON

EL SALVADOR, CENTRAL AMERICA

Sharat Kumar Roy

Chief Curator, Department of Geology

The 1917 eruption of Volcan Boqueron took place early in June
of that year. The event was reported at that time in the local
newspapers, La Prensa and Diario Latino. Later, it was described
by Emanuel Friedlander (1918), Karl Sapper (1926, pp. 243-246),
and others.

Unfortunately, none of these later authors saw the eruption.
Friedlander was in Zurich, Switzerland, and he based his account
partly on European newspaper reports and partly on official an-
nouncements that were made available to him by the Spanish
Consul in Zurich, Senor M. De Soto. Sapper, who returned to
El Salvador seven years after the eruption, states that most of the
information he has used for his article was furnished by Edward
Fischer, a naturalized Salvadorean, then residing at Santa Lucia,
a coffee plantation (finca), some nine kilometers west of the city of
San Salvador. Fischer was one of the eyewitnesses of the eruption.
His account of the event has been most useful in compiling this
account of one of the outstanding volcanic eruptions in Central
America during the present century.

An attempt has been made here to verify the discrepancies of
previous authors, especially with regard to time, date, and certain
eruptive phenomena, and to reconstruct the event accordingly. I,
too, was aided by my interview with Fischer, as was Sapper. The
reports of Jorg6 Larde (1917) have also been most helpful. Purely
descriptive material has largely been eliminated and certain salient
features arising out of the eruption, which have not hitherto been
dealt with and which I have observed in the field, have been noted.
It would seem that certain aspects of the causes and effects of a given

Library of Congress Catalog Card Number: 57-13311+
No. 824 363

364

FIELDIANA: GEOLOGY, VOLUME 10

Fig. 143. Map of Volcan San Salvador group, showing distribution of the
three units composing the group. Adapted from topographic map of Nueva San
Salvador, EI Salvador, Sheet 2357 III, AMS Series E 753.

eruption are more clearly understood and interpreted from the post-
eruptive field studies than at the time of the eruption.

Boqueron (or Quezaltepeque) is the largest and the next high-
est (1,885 meters) of the three units of a group of volcanoes, col-
lectively known as Volcan San Salvador. It is situated just west of
the capital of El Salvador, which bears the same name as the volcan.
Here it forms a conspicuous landmark and stands like a sentinel

ROY: VOLCAN BOQUER6n, EL SALVADOR

365

guarding the metropolis. Ironically, it is this so-called sentinel that
has been responsible for the repeated destruction of the city. Highly
localized earthquakes, usually with volcanic accompaniments, have
severely damaged or wrecked the city of San Salvador some thirteen
times (1538, 1575, 1593, 1656, 1659, 1707, 1719, 1793, 1814, 1839,
1850, 1873, and 1917) since its founding.

£.1 Jboqueron

P/cocho

9Sfm

Jca/t,

JOB

Fig. 144. Section through Boqueron and Picacho. The saddle is between
the two.

The other two units of the group are Picacho (1,967 meters) and
Amatepeque (1,397 meters); the latter is also called Jabali. The
distribution of these three units is shown on the sketch map (fig. 143) .
Picacho lies 3.3 km. northeast and Amatepeque 3.7 km. northwest
of Boqueron. The east flank of Boqueron and the west flank of
Picacho are separated by an elongated depression or saddle lying
between 1,400-1,500 meters; a road, which begins at Santa Tecla
(Nueva San Salvador) in the south and ends at San Juan in the
north, winds through it (see figs. 143, 144). Sapper (1925, p. 51)
mentions this saddle and states that it might be the remnants of
a crater, but he dismisses the subject with the remark (trans.):
"Only a prolonged investigation, for which I lacked time, could
settle the question." The saddle does not have any of the earmarks
of an old crater; it seems to be the natural lay of the land modified
by gradational processes — a feature which is found in many volcanic
areas between adjacent volcanoes. In most cases, saddles are broad
areas of union between spurs extending laterally from opposite
volcanoes or ranges of volcanoes. A striking example of this is Las
Brumas between Volcan Santa Ana and Cerro Verde in El Salvador.
The saddle between Boqueron and Picacho might well be the
meeting place of the spurs of the two. Subsequent deepening and

366 FIELDIANA: GEOLOGY, VOLUME 10

widening by erosion, or filling in by deposits of ashes and pyro-
clastics probably has taken place, partially obliterating the original
topography.

In contrast with the east face of Picacho, the west face is some-
what abruptly precipitous (fig. 145), unlike any natural development

Fig. 145. Volcans Picacho and Boqueron, facing south. Note abruptly
steep west face of Picacho.

of a volcanic cone; it suggests that this deviation from the normal
was caused by a succession of violent explosions subsequent to the
formation of the cone. The question naturally arises: What became
of the thousands of cubic tons of materials that were shattered by
the explosions? I have climbed to the summit of Picacho several
times from different directions and have seen individual blocks of
rocks on the slopes of the cone and around it but never in sufficient
quantity to constitute the total mass lost. It may well be that,
during the explosions, most of the rocks were broken into small
fragments, which have since decomposed into soil.

On February 27, 1951, I made my first trip to the crater of
Volcan Boqueron with Helmut Meyer-Abich, then a recently ap-
pointed Government Geologist of El Salvador, with offices at the
Instituto Tropical. I wish to take the opportunity here to extend
to him my sincere appreciation for his unfailing courtesy and
friendly co-operation, both in the field and in the laboratory.

As we reached the summit of Boqueron the gigantic crater burst
into our view. I had never before seen such a spectacular crater
(fig. 146), and the suddenness of its appearance, the stillness per-

l*S5m

Fig. 146 (right). Crater of Bo-
queron, containing a small cone of
vesicular lava with a crater of its

Fig. 147. Volcans Picacho and
Boqueron, facing east. The two
might have been one volcano, as
indicated by broken line.

367

368 FIELDIANA: GEOLOGY, VOLUME 10

vading the darkened hole with steep walls, some 550 meters deep,
took me aback. An interesting feature of the crater is that it is
nested, containing at the center of its floor another volcano, a small
cone of vesicular lava with a crater of its own.

It was a cloudless afternoon and from the east rim of Boqueron
I could clearly see the sheer west face of Picacho, the low saddle
between the two volcanoes, and much of the surrounding landscape
as far as Volcan San Vicente, some 60 km. away to the east. At
first I thought that Boqueron and Picacho had been one volcano
and that they had been connected in the manner shown in the
photograph (fig. 147) and later separated by a rapid series of ex-
plosions. Field investigations that followed, however, bared no
satisfactory evidence in support of this inference. On the contrary,
it became more evident that Boqueron, Picacho, and Amatepeque
came into existence separately.

Petrographically, the rocks of Boqueron and Picacho are not
quite the same; the latter is more andesitic and it contains relatively
more feldspar (80 per cent) and less pyroxene. With regard to the
relative age of these volcanoes, it is well nigh impossible to reach
a definite conclusion; neither their distribution nor their relationship
to one another, so far as structural superposition is concerned,
throws any light on the question. Dense growth of vegetation has
added to the difficulty of making detailed structural studies of the
basal portions of the volcanoes. Sapper (1925, p. 51) states that
Boqueron is the younger but furnishes no proof of his statement.
The fact that it has been periodically active since the latter part of the
sixteenth century (the most recent activity being in 1917) is a good
indication that it is younger and upholds Sapper's concept. In
a group of volcanoes, the one that is active or has manifested periodic
activity is likely to be the younger (Izalco in the Santa Ana group;
San Miguel in the San Miguel group).

Meyer-Abich (1956) postulates a rather ingenious hypothesis1 of
the origin of Boqueron and offers it in support of his conception
that Boqueron is the youngest member of the group. He states
(trans.) : "I have the impression that these peaks (i.e., Picacho and
Jabali) had formerly been independent volcanoes between which
there stretched an oval basin, 5-6 km. in diameter. The Boqueron
itself, that is, the central part of the mass, appears to be of more

1 The hypothesis and the brief comment that follows were inserted after this
paper had been completed.

ROY: VOLCAN BOQUERON, EL SALVADOR 369

recent age and to have been formed from the interior of the above-
mentioned basin. Evidence supporting this hypothesis is found in
a break in the north slope of the mass which runs from the north
foot of Picacho (at about 1,300 meters altitude) westward in the
form of an arc toward the west until it reaches to Jabali. The
course of this break can be followed more or less clearly. Above
the break the slope of Boqueron (between 1,300 and 1,500 meters
altitude) is 13 per cent (7.5° i, whereas below the break the in-
clination of the slope of the old mass is about double (15°)."

Crater of
ftoftieroH,

Jca i c

Fig. 148. Profile representing slope at 89: 17' 15* N. Lat. between
13 ° 46' 28' and 13° 44' 47' W. Long. Broken line shows original slope.

What is termed here a "break" occupies a very limited area,
only a fraction of the total circumference of the base of the cone.
The profile (fig. 148) representing the slope at 89° 17' 15" X. Lat.
between 13° 46' 28" and 13° 44' 47" W. Long, (precisely above
and below the "break" cited by Meyer-Abich i hardly offers a basis
for assuming that there existed a basin at the present site of Bo-
queron. The change in the angle of the slope is not as abrupt as
Meyer-Abich indicated. Moreover, such a change could result from
unequal distribution of ejectamenta over a sloping surface that was
originally gradual, or from superimposed lava flows that have
stopped short of covering the entire slope. In fact, the course of the
break may well be the continuation of the saddle, referred to earlier,
which gradually terminates a little west of San Juan, and can not
be construed as a mark of delimitation of a structural basin.

I have skirted the entire circumference of the rim of the crater
of Boqueron (a distance of nearly 5 km.) at an average altitude of
1,800 meters. There are many vantage points from which many
of the physical features of the volcano can be observed. The huge
cone is steep and is dissected radially into deep gorges, more pro-

370

FIELDIANA: GEOLOGY, VOLUME 10

.

Fig. 149. A crater within a crater,
the two having a common central
vent.

Fig. 150. A nest of four craters.
Eruptive vent has shifted.

nouncedly (to a mature topography) on the NE-SE side. It is also
on this side that the rim of the crater has been damaged either by
erosion or during past eruptions for a distance of about 1.2 km.,
thus interrupting the continuity of what was once a circular rim
(fig. 143).

Except after heavy rains the crater is easily reached in a car
equipped with four-wheel drive. The dirt road from Santa Tecla to
San Juan (fig. 143) forks at an altitude of 1,500 meters on the east
face of the volcano. The left branch of the fork then winds con-
tinuously uphill through several fincas to an open area at 1,800

ROY: VOLCAN BOQUER6N, EL SALVADOR 371

meters, where it ends. From here a brick and cinder path climbs
to a little "lookout" at the very edge of the crater. Narrow serpen-
tine trails lead down the almost perpendicular walls to the floor of
the crater, the center of which, as stated earlier, is occupied by
a beautifully symmetrical cinder cone (locally called Boqueroncito)
about 40 meters high, with a funnel-shaped crater. This cone was
formed during the eruption of 1917, giving rise to a nested crater
or a crater within a crater. Centrally placed nested craters have
a common central eruptive vent (fig. 149) as is the case here; in the
eccentrically placed craters, there is a shift in the position of the vent.
In Volcan Santa Ana, which has four eccentrically placed craters,
the eruptive center has moved at various times to the east (fig. 150)
in the direction of Lake Coatepeque.

The descent and the ascent of the main crater of Boqueron, which
is about 540 meters deep, take about two hours. Neither process
is hazardous, although both are arduous. They should not be made
after dusk, for the trails are narrow and rocky and in places run
over ledges of sheer cliffs. The floor of the crater, which has an
altitude of 1,347 meters, is covered with scoria ejected during and
after the formative period of the Boqueroncito.

The rocks of Boqueron, whether older or newer, differ but little.
All are basalt or basaltic andesite, composed chiefly of phenocrysts
of basic plagioclase (Ab4 An6) and varying amounts of pyroxene and
augite; the groundmass is consistently hyalopilitic.

When I first visited the crater, there was a placard (since re-
moved), giving pertinent information in English and Spanish about
the 1917 eruption (fig. 151). Obviously intended for visitors, it was
posted prominently just outside the "lookout," on a bit of level
ground within a few feet of the eastern rim of the crater. It read :

DATA CONCERNING SAN SALVADOR OR QUEZALTEPEC VOLCANO

Geographic Position:

Crater 13° 43' 55* N. Lat. and 89° 17' 20" W. Long. Greenwich. Height above
sea level: 1,887 m. Diameter of crater: 1 ' •> ks. Circumference: 4,712 m., ap-
proximately 5 ks. Depth of crater: approximately 500 m.

Lake in the Interior:

In the past there was a lake at the bottom of the crater 400 m. in diameter,
the water of which evaporated as a result of the eruption that took place on the
7th June 1917.

Eruption in 1917:

On Corpus Christi Thursday, 7th June, 1917, at six o'clock in the afternoon,
a series of violent earthquakes were the beginning of an eruption of the volcano

372

FIELDIANA: GEOLOGY, VOLUME 10

J DATA CONCERNING
0OR0S QUCZALTEPEC'VOLCAMO
^C POSITION m^mmm

I C«ATC»-1J* AV - -w. |

• i ■ -fVtw.tM'M OiA

■n'1"" loencc 4,n;« *»r>Bi

I ta»'C3 APPROXIMATELY SOO "

LAKE IN THE INTERI09

rMF OAST THtQf WAS A LAKt J

r 're «oom in oiAitnoTMc wau

Fig. 151. Placard (now removed) posted at rim of Boqueron crater, giving
information in English and Spanish about 1917 eruption.

which burst out on the NW slope ejecting a huge current of lava together with
dense clouds of gas and fumes. Little craters remaining on that side of the volcano
are called craterlets (boqueron citos). The eruption took place in the evening
at 8:00 o'clock.

Eruption at the Big Crater:

Three days after the eruption took place the lake at the bottom commenced
to boil, forming dense clouds of vapor. By the end of June the lake had evapo-
rated completely and formidable explosions followed from a cleft that appeared
at the bottom of the crater, gas and ashes being thrown up in tall columns every
five or ten minutes through the cleft. These explosions continued at intervals

ROY: VOLCAN BOQUER6N, EL SALVADOR

373

each greater in length until the first days of November of the same year. The
cone at the bottom of the crater was formed by the ashes and hardened by the
rains. THE COMMITTEE FOR THE BEAUTIFICATION OF ROADS,
SUBURBAN AREAS, PATHS AND PUBLIC PARKS.

The above contents conform essentially to those cited in the
literature of the event. There are a few discrepancies, most im-
portant of which is that the earthquake that heralded the eruption
took place on Wednesday, June 6, at 6:55 p.m., not on Corpus
Christi Day.

Fig. 152. City of San Salvador, showing damage caused by earthquakes that
accompanied 1917 eruption.

According to Mr. Fischer, now Director of the Zoological Garden
at San Salvador, the second shock caused the most damage to the
settlements surrounding Boqueron, including the city of San Sal-
vador (fig. 152). The photographs illustrating the evaporation of
the crater lake (figs. 153-158) were made by Mr. Fischer. I am very
thankful to him for giving me permission to reproduce them and for
furnishing me vital statistics of the memorable eruption.

Following the second shock (June 6, about 7:10 P.M.), a fissure
burst out on the northwest side of Boqueron, along which sprang up
a series of craterlets (fig. 163). From these poured out streams of
lava that flowed fanwise in the direction of Sitio del Nino. The
rapidly advancing lava soon engulfed a large territory, went past
the Quezaltepeque-Ciudad Arce railroad line (fig. 160) and abutted
against the older lava fields, one of which resulted from the eruption

Fig. 153. Lake in crater of Boqueron before 1917 eruption.

Fig. 154. Evaporation of crater lake.

374

Fig. 155. Emergence of a cone (Boqueroncito).

Fig. 156. Beginning of an eruption of Boqueroncito.

375

Fig. 157. Eruptive activity of Boqueroncito.

Fig. 158. Boqueroncito after eruption.

376

Fig. 159. Bottom of crater of Boqueroncito, 35 meters deep.

Fig. 160. Lava on Quezaltepeque-Ciudad Arce railroad line some 6 km.
north from seat of eruption.

377

378
I

FIELDIANA: GEOLOGY, VOLUME 10

Fig. 161. Craterlets, many of which have coalesced, giving off steam and
other gases, and exploding intermittently.

of 1659. On or about June 11, barely five days after the eruption had
begun, the lava ceased to flow, but the craterlets, some of which
have already coalesced, remained active until the middle of June
(fig. 161). By this time the typically aa lava had covered an area
of nearly 16 square kilometers and had built up a jagged terrain,
much of which is today without vegetation and difficult to walk over.
Fumarolic activity is still in progress in two of the craterlets situated
at the southernmost tip of the fissure. At the beginning, the more
active craterlets were formed between 1,350 and 1,400 meters; one
located farther down the slope at an elevation between 650 and 580
meters was the most violently active. Besides pouring out lava in
profusion, it manifested the Strombolian type of eruption, exploding
frequently with deafening reports like claps of thunder. J. Larde
named it El Tornador (the Thunderer, fig. 162; Larde, 1917).
Strong but highly localized earthquakes also accompanied the
eruptions of this craterlet.

It may be brought to attention here that west of this 1917
fissure and its craterlets, and bordering the northeast side of Volcan
Jabali, there is an older radial fissure with similar craterlets (fig. 143).
These two fissures, which are approximately of the same length,
about 8 kilometers, run parallel, and their craterlets, which occupy
about the same elevations, face one another. The fact that they
are so arranged may be accidental, but their presence along the

ROY: VOLCAN BOQUER6N, EL SALVADOR

379

lets.

Fig. 162. El Tornador (The Thunderer), one of the most active of the crater-

Fig. 163. Schematic diagram showing origin of craterlets along flank fis-

sure.

respective fissures bears evidence to the fact that craterlets are of
common occurrence along flank fissures. They owe their origin to
diagonal feeder pipes (fig. 163). The pipes are secondary cracks
(offshoots) formed during the initial eruption and later enlarged by
explosive activity. Absence of craterlets indicates that the explosive
activity through the diagonal pipes has been at a minimum; lava
simply has risen through them, pushed aside the overlying partially
congealed materials and rolled down the slope. This phenomenon
was observed at night during the 1955 eruption of Volcan Izalco,
El Salvador.

Prior to the eruption, there existed a lake (fig. 153) in the crater
of Boqueron (Sapper, 1925). It occupied an eccentric position with
a surface level at an elevation of about 1,420 meters above sea level,

380 FIELDIANA: GEOLOGY, VOLUME 10

which would indicate that it was 80 ± meters deep. That the lake
level was once higher is evidenced by the remnants of a terrace on
the northeast side of the crater at an altitude of 1,670 meters. The
dimensions of the lake as estimated in 1895 by Sapper (1925, p. 51)
were: N-S, 400 meters; E-W, 350 meters.

The water of the lake was apparently not potable; it was covered
with water-bloom composed of green algae and had an obnoxious
odor. An analysis by J. Puente (Sapper, 1925, p. 51) showed that
in addition to organic matter it contained sulphates and carbonates
of calcium and magnesium. The potability of the water is brought
to attention here purposely. Despite the difficulty of getting in and
out of the crater and the long and exhaustive trek to Santa Tecla,
the nearest trading town, there is a handful of men and women who
actually live in the crater proper the year around. Perhaps the
populace was greater when the lake existed. Since no water is to
be found anywhere within miles except during the rainy season
(June-October), these people must perforce use the lake water.
They eke out an existence inside the crater by raising flowers —
various species of Hydrangea, Lilium, Geranium, Aster, and Dian-
thus. These are grown on the slopes and ledges where the rocks
have decomposed to soil, which, like all volcanic soil, is remarkably
fertile. The making of charcoal from the vegetation in and around
the crater is another source of living. El Salvador has no natural
fuel such as oil or coal, in consequence of which the country has been
so denuded of first-growth trees that wood must be sought even in
the mouths of volcanoes.

On June 9-10, when the fissure eruption was still in progress but
waning, the lake began to boil, emitting clouds of vapor that all
but obscured the view of the crater (fig. 154). At intervals, steam
clouds rose at both ends of the lake, during which the lake level was
seen to undulate. Evidently, intermittent eruptions at the bottom
of the lake were taking place. The boiling continued until June 28,
when the water had completely evaporated. Assuming that the
lake was ellipsoid and the basin more conical than rounded and 80
meters deep it contained some three million cubic meters of water.
The total heat required for the vaporization of this volume of water
under atmospheric pressure at an altitude of 1,420 meters above
sea level would be approximately 73.17 X 10u BTU's.

Concurrent with the evaporation of the lake, a cleft on the dry
floor was observed, through which, at the beginning, spurted mud
and clots of lava accompanied by steam and other gases (fig. 155).

ROY: VOLCAN BOQUER6N, EL SALVADOR 381

Soon this changed to a full scale eruption; up came bombs, lapilli,
and other ejectamenta. One explosion followed another at intervals
of 5 or 10 minutes and columns of incandescent lava and gases shot
high into the air, almost to the rim of the main crater. Out of this
spectacular display there emerged at the center of the old crater
a perfectly shaped cone (Boqueroncito) of vesicular lava and ashes
with a funnel-shaped crater (about 35 meters deep) that could be
clearly seen between eruptions (figs. 158, 159). Thus formed a cone
within a cone, the two having a common central vent. In eight
days, between June 28 and July 5, the cone rose to a height of about
40 meters. Curiously, in spite of the fact that the eruptions from
the newly formed volcano continued until November of the same
year, including heavy ash eruptions during July, August, and
September, the height of the cone remained practically unchanged.
Apparently, the ejectamenta fell far and wide; little or none was left
on the cone itself for its growth and development. It could be
inferred that the crater eruption would have been far greater,
perhaps disastrous, had not the flank eruption preceded it. Much
of the stored energy and materials and the attending explosive
was released before reaching the main crater.

In places, the rocks of the little crater are still lukewarm. They
show little or no signs of weathering outwardly or in the interior.
The dominant minerals — olivine, feldspar (labradorite) and augite
— as seen in thin sections, appear quite fresh. Yet wisps of grass,
Andropogon rufa (Nees) Stapf, the widely distributed tropical
clubmoss, Lycopodium cernuum L., and the blue-green alga, Pityro-
gramma calomelanos (L.) Link, were observed to grow sparingly both
on the slopes of the cone and on the sides of the crater. Another
species of grass, Hyparrhenia rufa (Nees) Stapf, closely related to the
turkey-foot that covers the drier places in prairies and sand-plains
of our northern states, was relatively abundant on the floor of the
main crater.

At the present time, there is no activity of any kind in the
Boqueroncito. A single fumarole, feebly emitting steam and sul-
phurous gas, was observed on the northwest wall of the main crater.

382 FIELDIANA: GEOLOGY, VOLUME 10

REFERENCES

Friedlander, Emanuel

1918. tlber den vulkanischen Ausbruch in San Salvador. Zeitschr. f. Vul-
kanologie, Bd. IV, pp. 193-201.

1918. Weitere Nachrichten uber den Ausbruch in San Salvador. Zeitschr. f.
Vulkanologie, Bd. IV, p. 308.

Larde, Jorge

1917. El Quezaltepeque: Su eruption y terremoto de junio de 1917 (reimpreso
en "Tribuna Libre" del 4, abril de 1956). San Salvador.

Meyer-Abich, Helmut

1956. Los volcanes activos de Guatemala y El Salvador. Anales del Servicio
Geologico Nacional de El Salvador, Bull. 3, p. 67.

Sapper, Karl

1925. Los volcanes de la America Central, pp. 50-54. Halle.

1926. Die vulkanische Tatigkeit in Mittelamerika im 20 Jahrhundert. II.
Teil. Zeitschr. f. Vulkanologie, Bd. IX, Heft 5, pp. 243-246.