The Beginnings of the Phoenician Presence

in Southwestern Andalusia
With Findings from the Excavations at Cabezo de San Pedro (Huelva),
San Bartolomé (Almonte, Huelva),
Castillo de Doña Blanca (Puerto de Santa María, Cádiz),
and El Carambolo (Camas, Sevilla)

Diego Ruiz Mata

Posidonius’s account, passed on to us by Strabo (3.5.5), is often used to ex-

plain the initiation of the Phoenician presence in the southern Iberian Penin-
sula. During his stay in Cádiz, Posidonius heard an ancient legend that had
been preserved in the temple of Melkart. According to this legend, two other
trips, to Sexi (Almuñécar) and to Onoba (Huelva), had preceded the founding
of Gadir, the oldest Phoenician city in the West. According to Velleius Pater-
culus (Historiae Romanae1.2), the founding of Cádiz took place around 1104
b.c. However, this date has not been corroborated archaeologically, and the
most ancient remains date to the beginning of the 8th century b.c., at the ear-
liest. Due to archaeological excavations in various necropolises and towns, we
currently have a more accurate knowledge of the pre-Phoenician cultural envi-
ronment, the economic factors that motivated the Phoenician seafarers, and a
more accurate date for their arrival.
Around the year 900 b.c., or possibly earlier, we see a culture in western
Andalusia that we refer to as the Tartessian Final Bronze. This culture pos-
sessed an economic base that provided the principal attraction for enticing the
Phoenicians to their western overseas experience and their settlement on Ibe-
rian shores. Using only the data from pottery and its range throughout the area,
we can delimit the geographic boundaries of the Tartessian Final Bronze. The

Editor’s note: This article appeared in Spanish under the title “Aportación al análisis de los inicios
de la presencia fenicia en Andalucía sudoccidental, según las excavaciones del Cabezo de San
Pedro (Huelva), S. Bartolomé (Almonte, Huelva), Castillo de Doña Blanca (Puerto de Santa
María, Cádiz) y El Carambolo (Camas, Sevilla).” The article was published in Homenaje a Luis
Siret (1934–1984) (Seville: Consejería de Cultura, 1986) 537–56.

263
 264 Diego Ruiz Mata

results are shown by the map in fig. 1. The Tartessian settlements were located
in the Spanish provinces of Huelva, Sevilla, Córdoba, and Cádiz, especially on
the most level land. In addition to lowland fields and a broad front along the At-
lantic coast, this area contains the basin of the Guadalquivir River, which is
bounded by the foothills of the Sierra Morena on the north and the chain of the
Béticos on the south. Although these boundaries define the area of the Tartes-
sian Final Bronze, we also note influences that reach as far as the southeastern
Peninsula and Estremadura. These influences are a result of trading relation-
ships that most likely intensified under Phoenician stimulation beginning in the
middle of the 8th century b.c.
In the region of Huelva, the settlements were concentrated where the city
of Huelva stands today (on its various hills) and along the Río Tinto (or Red
River), up to its source and in the surrounding area of lowlands. Some settle-
ments were established as a result of metallurgic activity and trade. In this
way we can explain the settlements at Cerro de Salomón and Quebrantahue-
sos (in the region of Riotinto), and Niebla, as well as the market center that
must have developed in the hills of the city of Huelva. Farther to the east we
find the fortified towns of Tejada la Vieja (Escacena) and Mesa del Castillo
(Villalba del Alcor), the metal-working town of San Bartolomé (Almonte),
and probably El Rocío, on the shores of the marshland.
Several geographic features bound the lower Guadalquivir: the plateaus of
El Aljarafe and Los Alcores, la Campiña, and the foothills of the Sierra
Morena. Settlements were located on the eastern side of El Aljarafe (Escacena
1982) near the Guadalquivir, and it is also the eastern side of Los Alcores that
is most populated (Amores 1982). The Tartessian communities settled on low
hills between the two plateaus, perhaps because of the possibilities for farm-
ing and cattle-raising in the area. We must also mention the excellent route
that passed through here to the interior and to the metal-yielding region of the
Sierra Morena, suggested by the location of the town of Setefilla (Aubet Sem-
mler, Escacena, and Ruiz 1985, 135–140).
Until now, not much field work has been done in the area of Gadir (Cádiz).
The data come from Mesas de Asta, on the border of the marshes of the Guad-
alquivir, and from the recent excavations at Castillo de Doña Blanca. But in
surveys within the city limits of Puerto de Santa María and the western zone
of Jerez de la Frontera, several settlements have been found that fill the gap
that existed previously.
Likewise, little research has been done in Córdoba, but the known sites show
influences from the lower Guadalquivir going up along the course of the river,
as Colina de los Quemados inside the city of Córdoba itself suggests. Other
sites are linked to the countryside around Sevilla, as in the case of Ategua.
In the area where the Tartessian Final Bronze Age developed, not all the set-
tlements can be dated to the same period. We see, of course, some nuclei that

The Phoenician Presence in Southwestern Andalusia

 The Phoenician Presence in Southwestern Andalusia 265

Fig. 1. Map of the distribution of Tartessian and Phoenician sites.

„ Tartessian sites
∑ Trading posts—Phoenician colonies
W Sites with Tartessian materials

were inhabited since Phase I of the pre-Phoencician era, but the population
grew under Phoenican economic stimuli. For example, the region that we are
surveying between the Guadalete and Guadalquivir Rivers became more in-
tensely populated, as was to be expected, beginning in the middle or at the end
of the 8th century. According to its excavator, Jesús Fernández Jurado, the site
of Tejada la Vieja dates from the same period as Cerro Macareno (Pellicer, Es-
cacena, and Bendala 1983).
Nevertheless, during Final Bronze I, the basic economic structures were
forged that would shape the Tartessian economy and the Phoenician interest in
the area, although it is possible that these structures were not yet well devel-
oped. In addition to agriculture and cattle raising, we refer to silver metal-
lurgy. Silver slag has been found in the strata of the pre-Phoenician phase in
Huelva, and possibly also in Almonte.
266 Diego Ruiz Mata

The information from Posidonius and passed on by Strabo (3.5.5) about the
founding of Cádiz is now being confirmed objectively and can be explained in
terms of economic interests and colonial possibilities. Likewise, the date
given by Velleius Paterculus for the founding of Cádiz, and consequently for
the Phoenician presence in the southern Peninsula, continues to be clarified by
the ongoing field work.
The purpose of this article is to present data on the beginnings of the Phoe-
nician presence in the southwestern Peninsula based on the results of recent
excavations. Huelva can be identified as the Onoba mentioned by Strabo and
probably was the heart of the ancient Tartessos. San Bartolomé was a metal-
working town where silver was extracted from the 8th century b.c. to the be-
ginning of the 6th, and it was probably linked to Cádiz. Castillo de Doña
Blanca provides data about Cádiz, since it is close to that metropolis and acts
as a mirror image. El Carambolo, located on the plateau of El Aljarafe in the
province of Sevilla, provides evidence of the Phoenician presence in the lower
Guadalquivir, as seen in the considerable amount of red slip Phoenician pot-
tery coming from that site.

I. Huelva: Cabezo de San Pedro

The city of Huelva is located at the confluence of the Tinto and Odiel Riv-
ers, on the southern part of a peninsula formed by these two estuaries. It is at
the foot of a chain of hills that rise to the north and northwest, with elevations
of 50 meters on the average. On these hills (cabezos), or more specifically, on
the slopes of these hills, we find the area of greatest archaeological interest.
The early settlement, and even the later one, was spread over these hills,
all of which presented similar conditions for habitation. The data acquired
through several archaeological excavations hint at a nuclear settlement on the
slopes of Cabezo de San Pedro, as well as on the other hills. This pattern has
produced nuclei of stratigraphic sequences in several zones, but we have not
yet found an uninterrupted sequence from the first phase of occupation until
abandonment. It was not necessary, however, to concentrate the population in
only one location, since each hill provided identical conditions for human
habitation.
In this context we find a stretch of wall that represents the earliest structure
reflecting the Phoenician presence in Huelva (Ruiz Mata, Blázquez, Martín de
la Cruz 1981: 179ff.). This wall, whose function and date have at times been
debated, is located on the western slope of the hill of San Pedro, at the begin-
ning of a small stream bed that separates this hill from the hill of Cementerio
Viejo (which has since disappeared).
 The Phoenician Presence in Southwestern Andalusia 267

Plate 1. 8th-century Phoenician wall at Cabezo de San Pedro (Huelva). The central
supporting column in this wall was constructed with ashlars (1 to 1.2 m long and 0.5 m
wide), using the header-stretcher method.

The wall runs from east to west and has been excavated along a stretch of
approximately 10 meters, with a gap of almost a meter on its western end. Its
width is not uniform, reaching 2 meters in the center (where a column, or pil-
lar, of ashlars is located) and narrowing at its ends to 1.30 meters.
As for its construction, the wall contains a column of large ashlars, 1 to
1.20 meters long and 0.50 meters wide. Its courses alternate, using a header
and stretcher arrangement. The column supports two extending wings made
of rough slabs of slate, lightly squared, along with rubble to fill the cracks.
The order of construction appears to be the following: first, the column was
erected, taking advantage of the lowest part of the depression and serving as a
foundation; afterwards, the rough slabs of slate were placed against the col-
umn, without digging a foundation trench, but rather adapting them to a pre-
viously leveled surface. The southern back face of the wall simply leans
against the hill’s wall of tertiary loam. The function has not been clarified, but
most probably it served as a retaining wall to contain the earth and erosion re-
sulting from rainfall. It is less likely that it served as a defensive wall, and cer-
tainly it did not form part of an area divided into rooms.
We have already mentioned how the settlements were spread over the hills
of Huelva, reaching to the zone where the sea touched these promontories. It
268 Diego Ruiz Mata

Plate 2. The Río Tinto, which flows from the mining region of Riotinto to Huelva. The
river has a striking reddish hue.

seems that there was no layout of streets and blocks, but rather the settlements
were adapted nuclearly to the slopes of the hills of San Pedro, Cementerio
Viejo (which has since disappeared), Molino del Viento (lowered by modern
construction at some points), and La Esperanza.
According to the existing archaeological data, the oldest nucleus of settle-
ment corresponds principally to the hill of San Pedro (Blázquez, Ruiz Mata,
et al. 1979; Ruiz Mata, Blázquez, and Martín de la Cruz 1981), which con-
tains strata from the pre-Phoenician Final Bronze (9th century b.c.). From the
8th to the 6th centuries the settlements spread across the hills of Molino del
Viento and La Esperanza, as a result of economic and, therefore, population
growth in Huelva, based on bronze and silver metallurgy.
The excavations directed by Jesús Fernández Jurado at different points on
Molino del Viento confirm this organization of the habitat. They also confirm
that its chronology is more recent (beginning in the 8th century b.c.) than that
of the strata of San Pedro. These excavations have also shown that there are
frequent layers of erosion and mud brought down from higher areas in the
hills by torrential rains, a situation similar to the present day after periods of
intense rainfall.
The use of the slopes, the continuous danger of landslides of crumbly clay,
and the torrential rainfalls objectively confirm the purpose of the wall of San
 The Phoenician Presence in Southwestern Andalusia 269

Pedro as a way to contain erosion and to protect the inhabited areas below.
This is also the function of a wall using the same construction technique (col-
umns or pillars with extending walls of rubblework) discovered a few years
ago in Tyre (Elayi 1980; Bikai 1978).
Let us now look at the technique used to construct the wall in Tyre. J. Elayi
(1980) has published material on a type of Phoenician wall that he calls “pil-
lared,” which uses a technique similar to that of the wall of San Pedro. He dis-
tinguishes various types, but all have in common the use of columns of ashlars
and extending walls of rubble work. This technique originated in the North,
perhaps in Hittite construction during the Empire Age (1460–1190 b.c.), and
developed later in areas under Hittite influence. This technique reached its
high point on the central coast of Phoenicia between Joppa and Sarepta, al-
though it is seen along the whole coast. This technique was used not only for
public buildings, but also for dwellings, casemates, and certain temple zones.
In Tyre, it was used for a retaining wall.
Chronologically, the earliest walls are found at Ras Shamra in the 14th and
13th centuries b.c.; next are those of level IV of Megiddo (1000/800 b.c.),
level IX of Tyre (around 850 b.c.), those of Hazor (level VB, 760–740, reused
in level VA, 740–732, and level IV, from the end of the 8th century to the
middle of the 7th century b.c.), and Aroer (8th–7th centuries b.c.). At Motya
they are found in the 6th–5th centuries b.c., and at Nora and Carthage they are
found in the 3d–2d centuries b.c.
Area A.2.3 provides the best-defined stratigraphy for the dating of the wall
of San Pedro (Ruiz Mata, Blázquez, and Martín de la Cruz 1981). The wall lies
on top of levels of the pre-Phoenician Final Bronze I (Phase I.A). It was con-
structed during Phase I.B, and the strata that rest against its worked side (the
northern face) also belong to Final Bronze I. In other words, its construction
took place toward the end of Final Bronze I and represents one of the first
proofs of the Phoenician presence on the hills of Huelva. It was probably an
excellent way to establish friendly relations, and it provided a technical solu-
tion at the source of the small stream that separated the hills of San Pedro and
Cementerio Viejo to the constant problem of torrential rainfalls during the
spring and fall. It was built at the beginning of the 8th century b.c. (around
800), and the connection between this wall and those that were built on the
Phoenician coast is obvious. Stratigraphically this date appears certain, since
at the level of Area A.2.1 we found only one fragment made on the potter’s
wheel, a body sherd which appears to belong to a small oenochoe (Ruiz Mata,
Blázquez, and Martín de la Cruz 1981: figs. 41, 153). The rest of the material
is characteristic of the pre-Phoenician Final Bronze I. The retaining wall in
Tyre is dated to around 850 b.c. and thus probably predates that of San Pedro.
The dates obtained by Jesús Fernández Jurado in several excavations at
Molino del Viento in connection with the Phoenician presence also approach
270 Diego Ruiz Mata

Plate 3. Walls and square buttress tower of Tejada la Vieja.

the beginning of the 8th century b.c. And, although it was not found in context,
a fragment of Attic crater from the mid-Geometric II, dated between 800 and
760 b.c. and brought to this spot by Phoenician seafarers (Shefton 1982, 342),
indicates an early date for the initiation of Phoenician-Tartessian relations.

II. San Bartolomé (Almonte, Huelva)

The settlement of San Bartolomé (Ruiz Mata 1981; Fernández Jurado
1983) is located within the city limits of Almonte and corresponds to an indig-
enous community of metal workers who extracted silver during the Tartessian
period. The population was spread over four hillocks overlooking the stream
of San Bartolomé. This type of settlement pattern with dwellings in several
nuclei was normal for settlements of the Final Bronze. So far, more than thirty
huts from the 8th and 7th centuries have been excavated.
The location and the metal-related activities of San Bartolomé can be ex-
plained only in relation to mining centers and trade routes, since San Barto-
lomé was not a mining town itself (there are no mines in its immediate
vicinity, nor for geologic reasons could there have been). Instead San Barto-
lomé must have been a metallurgic town which received ore for refining, as
also occurred in the city of Huelva (Garrido and Orta 1982).
 The Phoenician Presence in Southwestern Andalusia 271

Plate 4. Tejada la Vieja: possible area for washing ore.

San Bartolomé is located on a route that began, perhaps, in the region of

Riotinto, although more probably in Aznalcóllar. The route reached Tejada la
Vieja, and went from there, via Manzanilla, to San Bartolomé and El Rocío,
on the banks of the Guadalquivir estuary. Remnants of slag from silver refin-
ing and other metallurgic activities are documented in Tejada la Vieja and El
Rocío. This route probably served the Cádiz metal trade. San Bartolomé’s ex-
istence was justified by the necessity of maintaining a route different from the
one that led to Huelva (the location of Tartessos) along the Río Tinto.
The analyses of the slag coming from the region of Riotinto and from San
Bartolomé provide data that do not contradict the results obtained by compar-
ing the ceramic typology and analyzing the ware. It has been calculated that
twenty million tons of slag resulted from mining the Riotinto area in ancient
times, of which four million tons probably came from the Tartessian period
(Pinedo Vara 1963, 25). When the slag from Riotinto was analyzed, the fol-
lowing results were obtained: Ag: 97%; Au: 0.10%; Bi: 0.42 %; Fe: 0.09%;
Pb: 0.14%; Cu: 2.25%. The results from the analyses performed on the ores
and slag of San Bartolomé (Ruiz Mata and Fernández Jurado) suggest the pos-
sibility that the ore at San Bartolomé did not come from Riotinto, but perhaps
from the mines of Aznalcóllar (province of Sevilla), based on the lead and
copper contents that differ notably from one site to the other. Riotinto provides
 272 Diego Ruiz Mata

slag that contains copper and little lead, while the slag from San Bartolomé has
a greater content of lead and less copper. Percentages similar to those at San
Bartolomé are provided by two samples of slag at Castillo de Doña Blanca
from strata of the 7th century b.c.
The composition of clays has been analyzed (Galván) by means of X-ray
diffraction, electron microscopy, and infrared spectroscopy on handmade and
red slip pottery from the towns of San Bartolomé (Almonte), Castillo de Doña
Blanca, San Pedro, Valencina de la Concepción, and Cerro de la Cabeza (San-
tiponce, Sevilla). Similarities have been noted between the wheel-made pot-
tery in San Bartolomé and in Castillo de Doña Blanca. The results show that
the wheel-made pottery from San Bartolomé and Cerro de la Cabeza does not
contain the same mineral components as the handmade pottery at these sites,
but rather that their composition matches the red slip pottery from Castillo de
Doña Blanca. The analyses seem to suggest imports of red slip pottery from
the area of Gadir to San Bartolomé and Cerro de la Cabeza.
In the strata of San Bartolomé, the earliest period corresponds to a horizon
of the Chalcolithic period, with foundations of circular huts and ceramic types
that are related to Huelva and the lower Guadalquivir. After a hiatus in which
we do not see Middle or Late Bronze materials, the site was reoccupied during
Final Bronze I, perhaps during the last part of this phase (second half of the
9th century b.c.), and it was inhabited until the end of the 7th century or be-
ginning of the 6th century b.c.
Huts XXXII/XXXIII and V provided the earliest material from the proto-
historic settlement of Final Bronze I in the period preceding the Phoenician
presence in the town. The ceramic types have been partially published (Ruiz
Mata 1981). However, in regard to types of cooking dishes (especially the
fabrication of the rims, the brown and grayish ware, and the gray and black
well-burnished surfaces with a metallic look), the most direct connections
may point toward the area of Huelva, the cultural nucleus for this community
of metal workers. Connections with the lower Guadalquivir are not entirely
absent, though, as suggested by fragments of large cups and cooking dishes
with red monochrome geometric motifs generally similar to those at El Ca-
rambolo and the lower Guadalquivir.
San Bartolomé has been extremely interesting because of the opportunity
to extensively excavate a habitat with separate dwellings dating to this period.
At times excavation is difficult in places where people have lived without in-
terruption and where thick layers of strata have accumulated. However, the
huts at San Bartolomé have a fragile structure and lack thick strata, suggesting
short, temporary occupations, noted also in the uniformity of ceramic types
that have provided a well-documented horizontal stratigraphy. In regard to the
theme that concerns us here, we can be precise about the first Phoenician re-
mains in the settlement, the absolute percentages (since each hut presupposes
a closed grouping), and the course of development following the first con-

spread is 1 pica long

 The Phoenician Presence in Southwestern Andalusia 273

1 2
10

11
4
3

12
5 6

8
9

Fig. 2. San Bartolomé (Almonte, Huelva).

tacts. We note, though, that San Bartolomé sprang up only because of silver
metallurgy and probable trade with Gadir, and it did not achieve a high degree
of orientalization during its existence. Rather it maintained its indigenousness
and its system of living in huts, without the urban development that other
towns experienced during this period in the region of Huelva and the lower
Guadalquivir. The interpretation of the material data also suggests indigenous
control of the metallurgic resources.
Another group of huts (XXXIV, XIV.A and XIV.B) provides later hand-
made materials. The cooking dishes, for example, lose the carination of their
exterior rim, as do the cups. The closed forms are substituted by types that
show evolved or new formal aspects. However, it is basically the same cul-
tural world, with a new look to its pottery. Materials made on the potter’s
wheel are very scarce and constitute a very small percentage when compared
to the indigenous pottery made by hand: six fragments from Hut XXXIV,
three from Hut XIV.A, and only one from XIV.B. The forms do not show
much variation: a bowl covered with a red slip (fig. 2.3), another bowl devoid
of decoration (fig. 2.1), a rim of a fragmented red slip plate with an undeter-
mined width, a small juglet with a short turned-out rim (fig. 2.5), a bottle neck
with a thick rim and carination on its lower edge (fig. 2.2), a body sherd from
a red slip oenochoe, a gray plate with a concave rim and burnished/polished
surfaces, two fragments of urns decorated with bands, and the mouth of an
amphora (fig. 2.4). Sherds of red slip plates are not very common, with a pref-
erence for bowls with a red slip on the interior or no decoration at all. During
this initial period the demand for pottery was not very great. In the next phase,
the gray pottery gains a much higher percentage. The presence of juglets and
oenochoes as exotic materials in this indigenous environment may show that
perhaps they were valued for their contents and used as gifts. This is probably
the significance of the amphora in Hut XIV.A.
 274 Diego Ruiz Mata

A later phase is distinguished by the appearance of a new type of hand-

made cooking bowl with a short and angular rim and by new styles of cups.
There is greater use of impressions made with fingersprints on the shoulders
of rough-surfaced pots and of decorations incised on the bell-shaped à char-
don pots and vases. Wheel-made pottery is now more common, though its
percentage is not high and it does not compete seriously with the usual indig-
enous types. Its presence is the result of more intense relationships rather than
daily living side by side, and the settlement maintains its arrangement of an-
cestral dwellings and its own pottery. In this phase the style of wheel-made
pottery is more varied. For example, the number of bowls decorated with an
interior red slip increases, and the gray pottery is quite a bit more frequent (fig.
2.6, 7). We now notice a gray bowl with a rim reinforced on the inside, which
was absent during the former phase. The fragments of amphorae (fig. 2.8, 10)
and urns (fig. 2.12) increase considerably. In general, the demand for Phoeni-
cian products is greater, and the ceramic style is more varied.
Phoenician interest in the settlement was probably irregular and exclusively
commercial, related to the silver work that was done there. It also appears that
San Bartolomé was a seasonal home for the indigenous inhabitants. We cannot
perceive even a trace of urbanism in the whole area excavated. The settlement
was abandoned while the indigenous pottery was still prevalent.
As for the Phoenician presence at the site, we must emphasize the scarcity
of sherds in the initial period and the absence of characteristic forms which
can be dated. The first period of occupation is indigenous, perhaps an ad-
vanced phase of Final Bronze I, and lacks materials made on the potter’s
wheel. During the evolution of the handmade pottery, the Phoenician imports
begin with the types mentioned above. There are no red slip plates (perhaps
they were not in demand), a phenomenon which also occurs at the nearby
town of Tejada la Vieja, according to our information from Jesús Fernández
Jurado (director of the excavations). Neither are there other forms which can
be dated more precisely. However, the open forms made on the wheel are
bowls (as in Tejada la Vieja), which are less common in the towns on the coast
and in Huelva itself. The indigenous forms, which are typologically related to
Huelva and the lower Guadalquivir, can be dated to the end of the 8th century
b.c. and the beginning of the 7th, which probably was the time of the Phoeni-
cian arrival in San Bartolomé. In regard to the origin of the wheel-made pot-
tery, we have indicated above how the composition of the clay shows the
same mineral components as that of Castillo de Doña Blanca, which probably
means that they originated in Gadir.

III. Castillo de Doña Blanca

Castillo de Doña Blanca is the name for a watchtower constructed in the
15th century a.d. and for a tell, more or less rectangular in shape, about 300

spread is 1 pica long

 The Phoenician Presence in Southwestern Andalusia 275

meters long on its east-west axis and 200 meters wide on its north-south axis.
The tell has an elevation of 34 meters above the present sea level, containing
from 8 to 10 meters of strictly archaeological strata.
The site belongs to the city of Puerto de Santa María (province of Cádiz)
and lies about three kilometers off the highway on the side road that leads to
Jerez de la Frontera through El Portal. It is located at the foot of the Sierra de
San Cristóbal, a range of hills with a smooth outline and low elevation, rarely
exceeding 100 meters, and on the border of an ancient estuary whose waters
once reached the city. At one time this estuary increased the size of the present
Bay of Cádiz up to the base of these hills. However, it has filled up gradually
with alluvial silt from the Guadalete River, which in ancient times flowed into
the sea near the site of Castillo de Doña Blanca. This site, therefore, has mean-
ing only as a port closely related to the Phoenician city of Gadir and the indig-
enous market.
The first description of the site comes from F. Ventura López (1923), who
was still able to see surface remains of streets, rooms, and a wall more than
three meters thick. On the necropolis he saw “Phoenician tombs, and tomb-
stones with Iberian and Tartessian characters.” F. de Ciria y Vergara (1935), a
local scholar, describes the site in similar terms. César Pemán (1941) was also
familiar with these ruins, suggesting that they could be some topographic
point cited in the ancient sources about Tartessos, perhaps the port of Tartes-
sos itself. A. Schulten (1943) could still see several vestiges of construction,
including the wall that surrounded the town, and he identified it with the port
of Menesteo, mentioned by Strabo, Ptolemy, and Marcianus.
In spite of the fact that there were references to its location in the scientific
bibliography, the site was never systematically excavated. The materials col-
lected from its surface, which L. Peña Basurto had referred to at times (1952),
were never published or evaluated culturally or chronologically. Only re-
cently, due to surveys performed by G. Chic García (1979), have we realized
its true significance as a pre-Roman town from the 9th–8th centuries to the
3d–2d centuries b.c. In 1979, the Subdirección General de Arqueología
granted us permission to administer the excavations, which continue up to the
present time.
The principal motivation for this excavation was the opportunity to re-
search a protohistoric site close to Cádiz (such as Doña Blanca) in order to
provide the data that we lack for Gadir concerning its cultural characteristics
and chronology. We have finished four seasons at the town in the southeast
and southwest corners (1979, 1981, 1982, and 1983), and one season on the
necropolis (1984).
The Phoenician presence in the town is the theme that interests us in this
article, but it is also appropriate to briefly present the cultural and chronologi-
cal development of the site as seen through the stratigraphy excavated up to
this point.
276 Diego Ruiz Mata

Chalcolithic Period
The earliest ruins in the area correspond to the Chalcolithic period. They
are spread around the base of the Sierra de San Cristóbal, and they occupy the
whole area of Doña Blanca and the area called “La Dehesa” to the east. Circu-
lar huts have been excavated that reach 2.5–3 meters in diameter, with foun-
dations of rubblework placed against the sides of the circular pit that had been
dug to a depth of 0.40–0.50 meters. These foundations probably supported
walls of mud-brick. Other smaller circular and elongated pits have been exca-
vated that were probably used for storerooms or toilets for the dwellings. The
pottery is of poor quality and consists of closed vessels, bowls of different
sizes and depths, and plates with almond-shaped or thickened rims, typologi-
cally matching that of the lower Guadalquivir and Huelva. Decorated pottery
is scarce, with the exception of a bowl or shallow plate that has black radial
lines painted on the polished interior surface.
In the southeast zone of the town of Doña Blanca, there is a stratum on top
of bedrock that may correspond to a later phase of the Chalcolithic period, and
from it we have gathered numerous sherds with a late Bell Beaker decoration
and form. The small space did not permit us to note the features of the dwellings.

Hiatus
Later superimpositions have not been found in La Dehesa. In the excava-
tions done in the town, the strata that correspond to the Chalcolithic period are
covered by a sterile layer of reddish clay 0.50/0.70 meters thick.

Final Bronze I
Within the town of Doña Blanca we have not detected clean strata that cor-
respond to this period. But to the northeast of the town, on the slopes of the Si-
erra de San Cristóbal and among the surface materials at La Dehesa, we have
found materials from Final Bronze I that surely came from the groups of huts
located in this area. We have found carinated cooking bowls, burnished deco-
rations, and fragments of large vessels that sometimes have a thick layer of
red ochre on the outside. We have also found stands (in one case, with a red
line decoration on the central support), rough-surfaced pots, and a fragment of
a large vessel with a burnished exterior and red geometric design, of the type
that has been called “Carambolo” (from the zone known as fondo de cabaña)
and that we prefer to call Guadalquivir I. These are the most characteristic ma-
terials from this period.

Final Bronze I–II

Commencing with this period (the first half of the 8th century b.c., perhaps
at the very beginning of the century), the founding of Doña Blanca took
 The Phoenician Presence in Southwestern Andalusia 277

place, evidenced by the first remains of rectangular rooms with red clay sur-
faces and walls of rubblework plastered with lime. The existing data show
that probably only the southern part of the site was occupied during this pe-
riod, which was the terrace nearest the water. It is here that we can observe the
initial contacts between the indigenous population and the Phoenicians that
had settled in Gadir.
The greatest percentage of pottery is fashioned by hand and corresponds to
that of an advanced stage of the Tartessian Final Bronze. The Phoenician pot-
tery is similar to the earliest pottery from Morro de Mezquitilla and Chorreras
in Málaga and constitutes the earliest evidence that we have of the initiation
of the Phoenican presence in Doña Blanca. We will deal further with this
topic below.

7th Century b.c.

This is a period of great urban activity during which the city was enlarged.
The excavated remains belong to dwellings with limestone structures, walls
of plastered mudbrick, and walls of well-executed rubblework, such as foun-
dations. During this century we note two or three superimposed architectural
structures, as well as greater stratigraphic depth.
We have excavated a stratum that can be dated to the beginning of the 7th
century b.c. From this point on and in a process of orientalization that must
have occurred very rapidly in this town, the largest percentage of pottery was
made on the potter’s wheel. The handmade pottery is less common and the
cooking vessels lose their distinctive carinations and rims. At the same time
the shallow bowls with pointed rims are abundant, with burnished motifs that
are new and more poorly executed. The wheel-made pottery already shows a
wide variety, and we will point out just a few examples. The plate rims, which
are scarce, have widths of 4–5 centimeters. The red slip carinated bowls are of
the classical type from Toscanos and Mogador. The lamps now have two
wicks and lack the flat bases of those in the older strata, and they are covered
by a red slip. The amphorae are of the de saco type with solid, thickened rims
on the interior. We now see a large number of urns with cylindrical necks and
bodies, and those with larger dimensions (or pithoi), with decorations of
bands, etc.
During the rest of this century, the percentage of wheel-made pottery in-
creases, while the handmade pottery is reduced to bowls with pointed rims
and some rough-surfaced pots. The rims of the red slip plates acquire a greater
width (6–7 centimeters), while the other wheel-made forms show hardly any
variation. Indeed, the impression is that the town of Doña Blanca, due to its
proximity and contacts with Gadir acquired a high degree of orientalization in
all of its aspects from the beginning of the 7th century b.c. on.
 278 Diego Ruiz Mata

6th Century b.c.

During this century, and more precisely at the middle, the wheel-made
types of pottery continue to evolve, and at the same time (as occurs also in the
colonies at Málaga and Huelva), importations from eastern Greece begin
(SOS and Corinthian amphorae and Ionian and Attic cups). During this cen-
tury we can point to a certain decline in construction and less stratigraphic
depth. Some pottery forms appear to be of Carthaginian origin. We begin to
glimpse ceramic types that are a prelude to the Turdetanic culture.

5th Century b.c.

This century witnesses a resurgence of the city, which acquires a layout of
streets and blocks. The ceramic types evolve toward the classic Turdetanic
forms, although certain orientalizing forms are maintained. It is at this time
that trading posts for the salting industry begin to function, and we note at
least twenty nuclei (of families?) on the coast between Puerto de Santa María
and Rota who are involved in this industry, providing materials from this pe-
riod. In Doña Blanca we can document the same types of amphorae that were
found in strata in the agora at Corinth before 448 b.c. (Williams II 1978: 17–
19), in a warehouse associated with the fishing industry. In addition, we see
amphorae from the Greek colony of Marseilles, from the beginning or the first
half of the 5th century b.c. Certain Attic products should probably be dated to
the middle of this century, such as the so-called Cástulo type cups.

Later Developments
At the end of the 5th century b.c. and beginning of the 4th, the Attic im-
ports increase, while the Turdetanic ceramic type remains fully developed.
During the middle and second half of the 4th century b.c., relations with
northern Africa are frequent, demonstrated by the appearance of amphorae
and a considerable quantity of imitation-Greek red pottery, like that found at
Kouass.
At the end of the 4th century b.c. and during the first half of the 3d century,
we see a gradual abandonment of the town, for reasons yet unknown. It is pos-
sible that the silt deposits in the Guadalete River made navigation difficult to
the port that had given the city its economic prosperity. Perhaps the popula-
tion was moving toward the coast of Puerto de Santa María-Rota, where a
prosperous salting industry was flourishing at a spot better suited for fishing
and salt.
Roman vestiges are scarce, suggesting the existence of a few homes. The
medieval remains of the 11th–13th centuries a.d. can be explained in the
same way.
Let us return to the theme that concerns us: the first period of the Phoeni-
cian presence in Doña Blanca. Areas 4 and 5 (excavated in 1983) have pro-

spread is 6 points long

 The Phoenician Presence in Southwestern Andalusia 279

vided the most data so far. Below we present the strata from Area 5 that
contain the earliest elements in that part of the town. The numbers given are
those of the complexes and were used during the course of the excavation.

2.035
This number corresponds to the stratum that lies on virgin soil and has pro-
vided material from the Chalcolithic period.

2.034
The earliest protohistoric materials from this area (the southeast corner
of the town) lie on a stratum of sterile reddish clay. This stratum provided
few fragments, mostly atypical, including body sherds of burnished cooking
vessels and rough-surfaced pots. Next to them were fragments of wheel-made
pottery, including a body sherd of an oenochoe, with grooves on the upper
section of the sherd.

2.033/2.032
These are strata with little depth and few sherds. They have yielded a rim
of a cooking vessel with a very pronounced carination on the outside and bur-
nished surfaces—a frequent type during Final Bronze I (Ruiz Mata). Also, we
found cups, rims of pots, and burnished decorations that had been reticulated
and incised. The wheel-made materials correspond to bases of plates and body
sherds of amphorae and red slip oenochoes.

2.031
This stratum has greater depth and more abundant materials. The rims of
cooking vessels have smooth carinations on the outside (fig. 3.1–5) and be-
long to a type that is found at a late stage of the Final Bronze—Phase I/II—in
San Bartolomé (Almonte, Huelva) (Ruiz Mata 1979, 3ff.). All of the bur-
nished decorations are reticulated with a narrow crosshatch and divided into
quadrants separated by burnished bands, as is frequent in the lower Guad-
alquivir. In addition, there are bowls (fig. 3.6, 7), à chardon cups (fig. 3.8, 9,
12), and rough-surfaced pots (fig. 3.11). The deep cups with an incised geo-
metric decoration continue. Among the wheel-made pottery we can point out
a body sherd of a carinated bowl (probably Samarian ware), made from a re-
fined clay and with a red slip and grooves, with no slip on the outside surface.

2.030
The largest percentage of the pottery is still handmade. The most charac-
teristic materials are cooking vessels with rims that are hardly distinguishable
(fig. 3.13–16), reticulated decorations with a narrow crosshatch, cups with
thin sides and concave rims (fig. 3.17), open vessels with bell-shaped necks
280 Diego Ruiz Mata

Fig. 3. Castillo de Doña Blanca (Puerto de Santa María, Cádiz).

 The Phoenician Presence in Southwestern Andalusia 281

(fig. 3.18–20), rough-surfaced pots, etc. Among the wheel-made material, we

found a saucer that was decorated on the inside with a broad black band on the
rim and with other narrower black/grayish bands below, over an orange back-
ground (fig. 3.22). Fragment 21 in figure 3 belongs to a similar saucer without
decorations. Other fragments are body sherds of urns and amphorae.

2.028
The ratio of handmade to wheel-made pottery appears more balanced. Ad-
ditionally, we note new features on the rims of the cooking vessels (fig. 3.23–
25), which tend to lose their external carination or to get thicker on the inside
(fig. 3.23). The reticulated, burnished decorations continue, and on some body
sherds we note a design with a wider crosshatch. The wheel-made materials
consist of a red slip plate rim (fig. 3.26), the side of a juglet (fig. 3:27), and
neck and body sherds of urns with a decoration of red and black bands.

2.027
A balanced ratio is maintained between handmade and wheel-made pottery.
The cooking vessels have rims that are similar to those of the previous stratum
(fig. 3.28–36), with hardly-noticeable carinations. However, as a novelty, the
rims of numbers 28, 31, and 32 tend to be concave. The cooking vessels with
incised decorations on the burnished exterior surfaces continue (fig. 3.37), as
do the rough-surfaced pots with short, concave necks and with rims or shoul-
ders that are sometimes decorated with fingerprints. Also handmade bowls and
urns with bell-shaped necks continue. The normal wheel-made pottery is
decorated with a red slip, and the pottery decorated with bands has become
scarce. The most frequent types are plates with a high-quality red slip (fig.
3.38–42); carinated bowls decorated in the same way, with very low sides (fig.
3.43, 44) or of a larger size with higher sides; neck and body sherds of oeno-
choes; amphorae with short necks (fig. 3: 46, 47) and prominent rims on the
outside (fig. 3.48); juglets (fig. 3.45); flat-based lamps without a slip; a few
body sherds of urns decorated with bands; and ostrich eggs, with remains of
ochre inside and a line decoration on the outside.

2.026
We see characteristics similar to those of Stratum 2.027. On the cooking
vessels we note types with evolved rims (fig. 4.1–8) and hardly any carina-
tion, or that become thicker on the inside, or that are concave, decorated with
medium and narrow burnished crosshatching. The cups with concave rims
continue (fig. 4: 10, 11), among which we should mention a cup with very thin
sides, well-burnished surfaces, and an interior decoration of wide burnished
bands (fig. 4.9). Other materials are pots with concave necks (fig. 4.12), á
chardon vessels (fig. 4.13), and sherds of cups with an incised decoration
282 Diego Ruiz Mata

Fig. 4. Castillo de Doña Blanca (Puerto de Santa María, Cádiz).

Fig. 5 (opposite). Castillo de Doña Blanca (Puerto de Santa María, Cádiz).

The Phoenician Presence in Southwestern Andalusia 283
284 Diego Ruiz Mata

(fig. 4:14). The wheel-made pottery does not differ from the earlier forms: a
red slip plate (fig. 4.15), a red slip carinated bowl with high sides (fig. 4.16),
and an amphora with a short, pointed rim (fig. 4.17).

2.024
The handmade cooking vessels continue their evolution, with a tendency to
lose their carinations (fig. 4.18–25). The same handmade and wheel-made
materials continue (fig. 3.33–41).

In Area 4 (fig. 5) we have also reached virgin rock. The lowest stratum cor-
responds to a horizon of the Chalcolithic period, and after a hiatus (a sterile
level of reddish clay), we find the protohistoric settlement. The handmade
cooking vessels have sides with slight carinations (fig. 5.1–8) and are deco-
rated with reticulated designs (fig. 5.12, 13). The rough-surfaced pots have
short, concave necks (fig. 5.9–11). The Phoenician wheel-produced material
shows the same repertoire as Area 5 (fig. 5.14–29). We should point out that
these strata are characterized by a low percentage of pottery that has been
painted with bands, in favor of red slip pottery.
In summary, these are the data for the first period of settlement and devel-
opment in the southeast sector of the town of Doña Blanca.
This zone, especially the strip closest to the ancient coastline, was occu-
pied during the Chalcolithic period by a population that was culturally similar
to that of the lower Guadalquivir and Huelva. The zone diminished in popula-
tion toward the end of the Bell Beaker period, as shown by the degenerate ce-
ramic forms and decorations. Probably, after a hiatus, a small community
belonging to the Tartessian Final Bronze was established (perhaps in the 9th
century b.c.), located on the far eastern end and on the sides of the Sierra de
San Cristóbal. Scattered remains have been found from this period, but they
are few in number, suggesting a small nucleus of population. It is evident that
the Phoenician presence in Gadir and the initiation of organized trading ac-
tivities attracted a more numerous population.
The spot where the town of Doña Blanca was to be established brought to-
gether optimum conditions, since it had an excellent natural port and was
close to the Guadalete River on the east, the best route toward the interior. To
the north is the Sierra de San Cristóbal, low in elevation and easily crossed,
and from there a route departed toward the estuaries at Mesas de Asta and
Ebora. After crossing the estuary of the Guadalquivir, one could reach El
Rocío and Almonte, where we have verified the existence of settlements that
refined silver. The island of Cádiz required support points on the coast that
were in contact with the lower Guadalquivir and the metal-producing zone,
but which avoided the route that led to Tartessos. For this reason, the settle-
ment at San Bartolomé (Almonte) was economically linked to Gadir. The for-
 The Phoenician Presence in Southwestern Andalusia 285

tified town of Tejada la Vieja (Escacena, Huelva), which Jesús Fernández

Jurado is now excavating, also seems to be part of this network. We should ad-
ditionally mention the surveys that are being performed between Puerto de
Santa María and the Guadalquivir River. These surveys are documenting a
route which linked Final Bronze I–II settlements containing orientalizing
materials. This route led toward the area of Ebora, where one could cross the
river and reach El Rocío and Almonte. Another obvious route is the Guadal-
quivir River itself.
Thus, the location of Doña Blanca can be justified. It was a trading center,
possibly controlled by Gadir, in an enclave that was favorable for reaching the
lower Guadalquivir, the Gaditanian mountains across the Guadalete River,
and the metal working areas. By situating itself on the borders of Tartessos,
Gadir could control agricultural and mining centers with the same economic
potential as Tartessos. This is probably the meaning of the text of Posidonius
when he mentions the two previous voyages (to Sexi and to Huelva) which
occurred before the founding of Gadir. In other words, the purpose of these
voyages was a reconnaissance of economic and colonial possibilities, infor-
mation about Tartessos, and the final founding of Cádiz, with all the advan-
tages already mentioned.
As for the period in which the settlement of Doña Blanca took place, we
presently rely on the data that we have described from Areas 4 and 5. The ear-
liest remains (related to the rectangular dwellings of rubblework) appear to be
found mostly on the southern side of the town, closer to the water. Perhaps
they indicate a small trading post with purely commercial purposes, which
later increased in size and urban organization during the 7th century b.c. The
earliest strata correspond to a late period of the Final Bronze, which we have
called Phase I–II at San Bartolomé and during which the first Phoenician ele-
ments appear in the indigenous environment. The handmade indigenous
cooking vessels (which are excellent guides for establishing the various
phases) show evolved forms, seen in the rims that have lost the prominent car-
inations that characterize the pre-Phoenician Final Bronze I in western An-
dalusia. The greatest percentage of pottery is indigenous, as can be expected.
But in little time, in a rapid process of rapprochement and assimilation, the
ratio becomes equal, as has been mentioned for Stratum 2.028.
Since the material is still being studied, it is impossible to be more precise.
However, in regard to the material produced on the potter’s wheel, it is pos-
sible to point out the almost exclusive use of red slip pottery and the scarcity
of vessels decorated with bands. As chronological indexes, we want to em-
phasize that the red slip plates have narrow rims, between 2.1 and 3.5 centi-
meters. When correlated with the data obtained from the trading posts in
Málaga (Schubart 1976), these plates can be dated to around the middle of the
8th century b.c. Nevertheless, these plates seem to be preceded by strata in
286 Diego Ruiz Mata

which wheel-made Phoenician materials have appeared, though this has not
been documented. Moreover, in another area that we excavated (Area 4), we
found plates with rims of around 2 centimeters at the bottom level, which at
Morro de Mezquitilla appear to indicate an earlier period (Schubart 1983)—
the first half of the 8th century b.c.
The Phoenician presence in Castillo de Doña Blanca probably began during
the first half of the 8th century b.c., at least according to the data provided so
far by the excavations. This indicates that Gadir was probably founded around
800 b.c., since we must keep in mind that the first Phoenician settlers did not
remain isolated for very long. They required coastal bases for the initiation and
expansion of their activities in order to meet their economic objectives.*

IV. The Town of El Carambolo

El Carambolo is a hill located on the eastern edge of the Aljarafe plateau,
91 meters above sea level and three kilometers from the city of Seville. Exca-
vations resulted from the accidental discovery of a treasure of gold pieces on
September 30, 1958 (Carriazo 1973). Excavations focused on the area where
the gold was found (an area known as fondo de cabaña ‘area of huts’) and
later on the hillside, which has been called poblado bajo ‘lower town’.
The earliest material from pre-Phoenician Final Bronze I was excavated in
the lower levels of fondo de cabaña, specifically in Strata IV and III. The up-
per strata (II and I) formed as fill on top of this area and therefore have little
relation to it, but correspond instead to the orientalizing period. We need to
make this point, since it has been frequently said that the four strata belong to
the fondo when in reality the only related strata are IV and III. In these two
strata are found materials characteristic of Final Bronze I, with the exception
of some wheel-made elements (Ruiz Mata 1981).
The later orientalizing strata are found in Strata II and I of fondo de cabaña
and also in poblado bajo, where an extended area of 375 square meters was
opened. In summary, four levels have been distinguished in poblado bajo (A,
B, C, D, according to the excavator) that correspond to the remains of super-
imposed rooms. It must be pointed out that sketches of the stratigraphies and
the location of materials in their corresponding strata are lacking. The difficul-
ties are at times insurmountable for an accurate review. In addition, the sherds
were identified with adhesive labels, which of course have fallen off. Paloma
Cabrera Bonet has sketched the red slip material in the Archaeological Mu-
seum of Seville and has kindly given us permission to republish it.

Editor’s note: See also the author’s update on Castillo de Doña Blanca in his other article in
this volume, pp. 170–198.
 The Phoenician Presence in Southwestern Andalusia 287

Fig. 6. Typology of the rims of handmade cooking dishes from poblado bajo in El
Carambolo.

The pottery from poblado bajo has been analyzed in a very general and
succinct way, and only on a few occasions are the strata from which it comes
identified. The data we have can be summarized by the following points:
(1) Four levels of superimposed construction have been identified, and the
earliest (Level D), on top of virgin soil, is associated with indigenous hand-
made material and Phoenician wheel-made pottery, probably imported to the
town.
(2) The painted monochrome pottery of the “Guadalquivir I” type is prac-
tically absent, except for three fragments that preserve vestiges of paint and
could be from a later style. The types that characterize fondo de cabaña, such
as cooking vessels, cups, stands, etc., were no longer in use.
(3) Handmade pottery still persists (abundantly in Level IV), but only the
cooking vessels and bowls with burnished decorations are mentioned and
sketched. All together, 340 fragments of cooking vessels were collected.
These are described in groups: 111 come from Level IV, 21 from Level III, 5
from II, and 9 from I. The cooking vessels have profiles and characteristics
which differ from the earliest in fondo de cabaña.
(4) According to the excavator, the wheel-made pottery appears to be very
abundant, but he does not indicate percentages or types for the various levels.
He mentions gray pottery, amphorae (that supposedly are half of all the frag-
ments excavated), pottery decorated with bands, and red slip pottery. The
great quantity of amphorae (50 percent of the wheel-made fragments) sug-
gests the importance of the spot for commercial interests.
The best-known form is the handmade cooking vessel (fig. 6), and most of
these vessels come from Level IV. The rims and surface finishes are distinct
from the cooking vessels in fondo de cabaña. Types I and II with concave
rims are the most numerous and can be dated to 725–700 b.c., as hut XIV.B of
288 Diego Ruiz Mata

6 7

10

11 12

13

14

15

16 17

Fig. 7. Red slip plates from El Carambolo.

 The Phoenician Presence in Southwestern Andalusia 289

San Bartolomé suggests (Ruiz Mata 1982) and to 650–625 b.c., according to
the chronology that Castillo de Doña Blanca has provided.
Type IV, with a short, angled rim, is not as frequent and is used more in the
region of Huelva, where it is dated to the first half of the 7th century b.c. The
stratigraphic position of Type V appears to be more recent than that of Types
I and II. This is also true for the Type IX bowl, which in poblado bajo appears
beginning at Level III. In summary, Level IV can be dated from the end of the
8th century b.c. until a little after the middle of the 7th.
Rims belonging to the transition phase (Final Bronze I/II) have not been
found among the cooking vessels at poblado bajo. However, they have been
found at Castillo de Doña Blanca and San Bartolomé, which is the context in
which the first signs of the Phoenicians appear. The oldest handmade materi-
als at poblado bajo (Level IV) start at the end of the 8th century b.c. and be-
ginning of the 7th, dates which are suggested also by the red slip pottery.
We have already mentioned that in the publication of these materials nei-
ther the strata nor the phases to which these materials belong were specified
and that the majority have lost the adhesive labels that recorded the data.
However, it is possible to correlate the type and quality of their slips with ma-
terials that have better stratigraphic evidence at other sites. The plates are rela-
tively numerous and can be compared by typology with those at the town of
Doña Blanca and at the trading posts of Málaga. Figs. 7 and 8 organize the
plates according to the width of the rims, and fig. 9 summarizes the data.
Narrow plate rims not exceeding 3.5 centimeters (those between 1.5 and
2.9 centimeters) are present on seven occasions (fig. 7.1–7). We know that
plate 5 (2.5 cm.) comes from Level D (or Stratum IV) and that plate 1 (1.5
cm.) comes from Level C (III). Both were found 3 meters down, according to
the labels. It is very possible that both fragments belong to the same stratum,
given the ambiguity with which the levels are described. The ware is orange
(only on two occasions do we see an orange core, fragments 5 and 7) with
very fine grits. The slips, of quite good quality (except on fragments 3 and 4,
which have very diluted slips), are thick and form a consistent film, and they
are brown or brownish-red in color. The slips cover the whole inside surface
of the plates; on the outside, the slips reach somewhat lower than the width of
the rim on fragments 5, 6, and 7, and they reach only to the upper area of the
rim on fragments 1 and 4.
The characteristics of the slips closely resemble those at the town of Doña
Blanca in levels that are older or contemporary, and the pottery at El Caram-
bolo probably comes from that region (either Doña Blanca or Cádiz). During
the initial phase of Doña Blanca, the width of the rims is between 1.8/2.0 and
3.5 centimeters. However, the handmade cooking vessels at Doña Blanca are
transition types (Phase I/II) from the middle of the 8th century b.c., with less
pronounced carination of the rims. The cooking vessels from poblado bajo (in
290 Diego Ruiz Mata

10

11

12 13

14

Fig. 8. Red slip plates from El Carambolo.

 The Phoenician Presence in Southwestern Andalusia 291

Level IV/D) are more evolved. When we correlate these cooking vessels with
those from other sites in southwestern Andalusia, they can be dated from the
end of the 8th century b.c. until a little after the middle of the 7th.
The majority of the plates have rims between 3.5 and 5.0 centimeters (fig.
7.8–17 and 8.1–8). The plates with rims between 3.5 and 4.0 centimeters are
covered with high-quality slips which extend over the whole interior and a
large part of the exterior surface, down to a point below the edge of the rim.
The plates that have rims of 5 centimeters or more are inferior in quality. The
slips are dark, light, or reddish-brown in color.
Another group of plates has rims between 5.5 and 8 centimeters wide (fig.
8.9–14), and they also are covered with dark, light, or reddish brown slips that
are uneven and faded in most cases. Keeping in mind the lack of stratigraphic
documentation for the materials, we can assume that the plates show a typol-
ogy similar to the evolution observed in the trading posts and necropolises of
Málaga and Granada and in the closer town of Castillo de Doña Blanca. Thus,
we can date the materials from the end of the 8th century b.c. to the beginning
of the 6th.
Other open forms (also without stratigraphic information) are the carinated
red slip bowls that show different tendencies in the formation of their rims
(fig. 10.1–8), documented, for example, in Toscanos (Schubart and Linde-
mann 1984) and in Castillo de Doña Blanca. Fragments 3, 4, and 5 (fig. 10)
have open rims with pointed edges and high carinations and are covered with
high-quality slips. Another type shows an almost vertical rim (fig. 10.1), cov-
ered with a light red slip. Fragment 3, in the same figure, has a vertical rim that
is shorter and pointed, with an intensely burnished light brown slip. Bowls 7
and 8 (fig. 10) have a smaller diameter; fragment 7 has a vertical rim and a low
carination, and fragment 8 has an extremely concave rim. Both bowls are cov-
ered with a high-quality brown slip.
Other open forms are the bowls with concave rims and high carinations
(fig. 11.1–4). These bowls are characteristically covered with very faded red-
dish brown and dark brown slips, in contrast to the consistent, thick slips of
the carinated bowls. This type is similar to the forms coming from strata of the
7th century b.c. in Doña Blanca.
The urns have a cylindrical, vertical neck (fig. 11.5, 7) with broad bands of
red slip on the exterior surface, or a short neck inclined slightly inward (fig.
11.6). The body tends toward a spherical shape. A geminate handle takes up
almost the whole neck and sits on top of the upper area of the body. Only the
edge of the rim is decorated with a covering of dark red slip. Fragment 8
(fig. 11) has a central raised band around its neck and a more prominent rim,
and it is decorated with a very faded red slip.
Finally we should mention that body sherds of oenochoes (fig. 11.9–11),
with grooves on the upper section and with burnished reddish-brown slips of
292 Diego Ruiz Mata

3 2

Fig. 9. Widths of rims of

the red slip plates from El 8
Carambolo (black circles) 7
and Castillo de Doña
Blanca (white circles).
9

10

11

Fig. 10 (right). Red slip bowls and

plates from El Carambolo.

Fig. 11 (opposite page). Phoenician bowls, urns, and oenochoes from El Carambolo.

high quality, are included in this collection of Phoenician pottery from

poblado bajo. The inside surface of fragment 10 is covered with a very faded
violet paint. Fragment 12 comes from the neck of a mushroom-mouth oeno-
choe, which is covered with a burnished cream slip applied vertically and a
red slip on the upper section of the mouth. Fragment 13 belongs to an oeno-
choe with a trefoil mouth and is covered with an uneven light brown slip.
 The Phoenician Presence in Southwestern Andalusia 293

5 6

8
7

12

10

13

11
294 Diego Ruiz Mata

The data provided by the excavator do not help much when we try to reach
a conclusion about the stratigraphic position of the pottery. We can give only
a global appraisal of the red slip ceramic type.
As for the plates, we see the same typology that is already well known at
the trading posts of Málaga and the current excavations at Castillo de Doña
Blanca. Fig. 9 illustrates the width of the rim for this type of plate. The black
circles show the widths from poblado bajo and the white circles show the
widths from the earliest levels of Doña Blanca, which we have discussed ear-
lier in this article. We see plates with narrow rims (up to 3.3–3.4 centimeters),
but the majority have widths between 3.5 and 5.0 centimeters. When we cor-
relate these data with those from the trading post at Toscanos, it appears that
rims with widths of less than 3.5 centimeters (Schubart 1976) can be dated to
the second half of the 8th century b.c. We can date the rim width of the im-
ports entering El Carambolo to around the transition from the 8th to the 7th
centuries b.c. or to the beginning of the 7th. We mentioned earlier that in
Doña Blanca the narrow-rimmed plates are found with handmade cooking
vessels that are lightly carinated on the outside, while in poblado bajo these
plates are found with cooking vessels whose rims are more concave and that
seem a little later. Perhaps the beginning of the imports should be placed in the
second half of the 8th century b.c., and probably closer to its end. The type
of ware and slip is similar to that at Castillo de Doña Blanca, which should be
a faithful reflection of Cádiz due to the proximity of the two sites. Once the
Phoenicians were established in Cádiz, probably not much time passed before
they sailed up the Guadalquivir. One of their points of contact must have been
the town of El Carambolo.
The remaining forms find strict parallels in the trading posts of Málaga and
Castillo de Doña Blanca. Carinated bowls like 1, 2, and 4 of figure 10 are
found in strata at Doña Blanca dated to the second half of the 8th century.
Number 3 appears a little later, during the first half of the 7th century b.c. The
bowls with concave rims have the same dates.
In summary, and keeping in mind the necessary reservations in this case,
the Phoenician presence in El Carambolo probably began in the latter part of
the second half of the 8th century b.c. and intensified at the beginning of the
7th. The Phoenician economic interest in this place must have been great.
Thus it is very interesting that the excavator mentions that almost half the pot-
tery that was found came from amphorae.

V. Conclusions
(1) Before the arrival of the Phoenicians in western Andalusia, a complex
nucleus of population (of unknown origin) had taken shape, principally in the
Basin of the Guadalquivir and Huelva. The cultural horizon known as “Tar-
 The Phoenician Presence in Southwestern Andalusia 295

tessian” was composed of their small settlements and economic activities.

During this period we see signs of metal working, with the goal of extracting
silver, in the strata of Phase I at Cabezo de San Pedro and on the floors of huts
in San Bartolomé. Thus the Tartessian economic activity that provided the ba-
sic attraction for the Semitic traders is clarified. However, within this homo-
geneous cultural horizon, we can evaluate the differentiating characteristics of
certain ceramic forms and of the burnished and painted decorations. This al-
lows us to identify at least two large nuclei, one around Huelva and the other
in the lower Guadalquivir.
The stratigraphic results and analyses of slags and clays suggest signs of
pre-Phoenician silver metallurgy, with considerable development under Phoe-
nician influence. They also suggest the existence of at least two foci of metal
extraction, the zone of the Río Tinto reaching down to Huelva, and the zone of
Aznalcóllar, which traded directly with the Phoenician colony of Gadir, avoid-
ing the former route. This hypothesis is also suggested by the analysis of slag
and of the red slip pottery from both zones, linking Gadir with San Bartolomé.
(2) When we combine these data with the information from Posidonius and
passed on to us by Strabo (3.5.5), the purpose of the founding of Gadir be-
comes clearer. According to these accounts, the first attempts at Phoenician
settlements on the Iberian Peninsula took place at Sexi (identified with Almu-
ñécar) and later, after passing through the Gates of Hercules, at Onoba (prob-
ably Huelva). The founding of Gadir occurred on the third trip. It is clear that
the purpose of the first voyage was to survey and explore the possibilities for
trade and settlement. On the second try the Phoenicians reached Huelva, a
center where silver-related activity probably began during the pre-Phoenician
era and where it would not have been easy for the Phoenicians to initiate an
advantageous trade during that initial period. However, it is obvious that rela-
tions began around this time, as suggested by the construction of the wall of
San Pedro and the red slip ceramic types from the first half of the 8th century
b.c. It is also probable that the Phoenicians who settled in Huelva had little to
do commercially with those who founded Gadir, as seems to be indicated by
the separate development of the pottery in Huelva. This pottery does not par-
allel the ceramic assemblage from Gadir, which we are now beginning to rec-
ognize because of the excavations at Castillo de Doña Blanca. In this regard,
Schubart (1976) noted differences in the typological evolution of plates from
the trading posts at Málaga and Huelva, differences that exist also in other
forms. Whatever the case, the Phoenicians finally founded Gadir, and from
there used a different route for transporting metal, a route that reached Az-
nalcóllar by way of El Rocío, Almonte, and Tejada la Vieja. This means that
Gadir avoided the route of the Río Tinto to Huelva, which was under indige-
nous control. In fact, there are no early Phoenician settlements on this section
of the Atlantic coast except for Gadir. Gadir was rather far from the centers of
296 Diego Ruiz Mata

production in Huelva, but had easy and perhaps politically favorable access to
Aznalcóllar and the rich agricultural and cattle-raising regions of the Guadal-
quivir. All of this probably took place during the first half of the 8th century
b.c., judging by the recent findings.
(3) We have analyzed certain material aspects of the Phoenician presence
in western Andalusia. Huelva has not yet provided clean strata of occupation
from the first period of the Phoenician presence. This is not because of lack of
excavation, but perhaps because of the system of nuclear settlements on hills,
the current systematic destruction that has erased a large part of the early evi-
dence, and the water table that prevents digging to greater depths. Neverthe-
less, the excavations at Cabezo de San Pedro provide elements that date the
Phoenician presence in Huelva to the first half of the 8th century b.c. and even
to the beginning of the century. This can be deduced from the retaining wall
itself, its stratigraphy, and the fragment of an Attic crater from the mid-
Geometric II, dated between 800 and 760 b.c.
Castillo de Doña Blanca has provided more pottery from this first period.
Because of its proximity to Cádiz (Gadir) and its location on the ancient coast,
Doña Blanca may reflect the early period of Cádiz, where early strata are still
unexcavated. The results of the findings at Castillo de Doña Blanca show that
the Phoenician presence in the bay of Cádiz can be dated, as in Huelva, to the
first half of the 8th century b.c.
The town of San Bartolomé was of great interest to us because of its silver
extracting activity, its clearly indigenous culture, its location in an area where
there were no mines (though surely a trade route), and its probable trading re-
lationships with Gadir. We have excavated a significant sample of about thirty
huts, which contain predominantly indigenous materials. However, these huts
also contain wheel-made pottery as a product of commercial contact, begin-
ning perhaps in the second half of the 8th century b.c. at the earliest.
Finally, El Carambolo has provided Phoenician materials, and its location
next to the ancient mouth of the Guadalquivir makes it an interesting spot to
evaluate the Phoenician presence in the rich cattle-raising and agricultural
centers of the lower Guadalquivir. We have pointed out the problems that re-
sult from archaeological work that lacks proper stratigraphic records. How-
ever, the typological correlations of the red slip pottery from El Carambolo
with the pottery from Doña Blanca and Málaga show that the Phoenicians
probably sailed to this area beginning at the end of the 8th century, and their
presence increased beginning in the first decades of the 7th century b.c.

Bibliography
Amores Carredano, F.
1982 Carta arqueológica de Los Alcores (Sevilla). Series 1, no. 22. Seville: Dipu-
tación Provincial de Sevilla, Sección Historia.
 The Phoenician Presence in Southwestern Andalusia 297

Aubet Semmler, M. E.; Escacena, J. L.; and Ruiz, M. M.

1983 La Mesa de Setefilla. Lora del Río (Sevilla). Campaña de 1979. EAE 122.
Bikai, P. M.
1978 The Pottery of Tyre. Warminster: Aris & Phillips.
Blázquez, J. M.; Ruiz Mata, D.; et al.
1979 Excavaciones en el Cabezo de San Pedro (Huelva): Campaña de 1977. EAE
102.
Carriazo, J. de M.
1973 Tartesos y El Carambolo. Madrid.
Chic García, G.
1979 Gades y la desembocadura del Guadalquivir. Gades 3: 7–23.
Ciria y Vergara, F.
1935 Tartesos. Puerto de Santa María (typed text).
Elayi, J.
1980 Remarques sur un type de mur phénicien. RSF 8.2: 165–80.
Escacena, J. L.
1982 Antiguas vías de comunicación en el Bajo Guadalquivir. Gades 9: 129ff.
Fernández Jurado, J.
1983 San Bartolomé de Almonte: yacimiento metalúrgico de época tartésica. Re-
vista de Arqueología 26: 24–33.
Galván, V.
1986 Análisis de pastas cerámicas. Chapter 29 in El yacimiento metalúrgico de
época tartésica de San Bartolomé de Almonte (Huelva), eds. D. Ruiz Mata
and J. Fernández Jurado. Huelva Arqueológica 8.
Garrido, J. P., and Orta, E.
1978 Excavaciones en la necrópolis de “La Joya.” Huelva. II. (3.a, 4.a, y 5.a
Campañas). EAE 96.
Pemán, C.
1941 Nuevas contribuciones al estudio del problema de Tartesos. AEArq 42: 177–
87.
Pellicer, M.; Escacena, J. L.; and Bendala, M.
1983 El Cerro Macareno. EAE 124.
Peña Basurto, L.
1952 Tartesos, la ciudad más antigua de Occidente, descubierta. Revista Munibe.
Pinedo, I.
1963 Piritas de Huelva. Madrid: Summa.
Ruiz Mata, D.; Blázquez, J. M.; and Martín de la Cruz, J. C.
1981 Excavaciones en el Cabezo de San Pedro (Huelva). Campaña de 1978.
Huelva Arqueológica 5: 149–316.
Ruiz Mata, D.
1979 El Bronce Final – Fase Inicial – en Andalucía Occidental: Ensayo de defini-
ción de sus cerámicas.” AEArq 52: 3–19.
1981 El poblado metalúrgico de época tartésica de San Bartolomé (Almonte,
Huelva). MM 22: 150–70.
Ruiz Mata, D., and Fernández Jurado, J.
1985 La metalurgia de la plata en época tartésica en Huelva. Pyrenae 21: 23–44.
298 Diego Ruiz Mata

1986 El yacimiento metalúrgico de época tartésica de San Bartolomé de Almonte

(Huelva). Huelva Arqueológica 8.
Schubart, H.
1976 Westphönizische Teller. RSF 4: 179–96.
1983 Morro de Mezquitilla. Vorbericht über die Grabungskampagne 1981 auf
dem Siedlungshügel an der Algarrobo-Múndung. MM 23: 33–45.
Schubart, H., and Maass-Lindemann, G.
1984 Toscanos, el asentamiento fenicio occidental en la desembocadura del río
Vélez. Excavación 1971. NAHisp 18: 39–210.
Schulten, A.
1943 Forschungen in Spanien. AA 58: 19–62.
Shefton, B. B.
1982 Greeks and Greek Imports in the South of the Iberian Peninsula. The archae-
ological evidence. Phönizier in Westen, Madrider Beiträge 8: 337ff.
Ventura López, F.
1923 Articles published in the newspaper El Guadalete (Apr. 8, 1923; Apr. 10,
1923; Apr. 11, 1923; Apr. 15, 1923; Apr. 24, 1923; Apr. 25, 1923, Apr. 27,
1923; May 3, 1923; May 8,1923; May 18, 1923; May 20, 1923; Aug. 12,
1923; Dec. 2, 1923; Dec. 5, 1923; Dec. 7, 1923; Dec. 11, 1923).
Williams II, C. K.
1978 Corinth, 1977: Forum Southwest. Hesperia 47.1: 1ff.