• 1 BORJA BARRERA, F.; BARRAL MUÑOZ, M. A. Evolución histórica de la llanura aluvial del Guadalquivir y (...)

6Finally, considering the important role that geoarchaeology can play in this particular study, there has been a review of specific literature and conversations have been maintained with the authors. May it be worthwhile to quote the statement made by Francisco Borja y Ángeles Barral1:

• 2 JOAQUÍN CORTES, José et al. Planos de Sevilla. Colección Histórica (1771-1918). Sevilla: Servicio d (...)
• 3 COLLANTES DE TERÁN Y DELORME, Francisco. Contribución al estudio de la topografía sevillana en la A (...)

7To this purpose I have consulted the test drillings corresponding to the mentioned archaeological surveys, as well as a few others conducted on the occasion of different geotechnical studies. These last ones have been kindly provided by the company Vorsevi S.A. Concerning the topographical data on the area, there was no access to a full-scale, up-to-date survey, though the water supply company EMSAESA could provide of a series of tacheometric points belonging to the area of study. As these did not produce wide enough a framework as to recreate the area´s context, those historical maps of Seville containing topographical information were consulted2. The topography annexed by Collantes de Terán Delorme to his 1977 study3 has also been retrieved and processed.

• 4 BARRAL MUÑOZ, Mª Ángeles. Estudio geoarqueológico de la Ciudad de Sevilla : Antropización y reconst (...)

8This study concluded before the release of the book “Estudio geoarqueológico de la Ciudad de Sevilla: Antropización y reconstrucción Paleográfica durante el Holoceno reciente” by Barral Muñoz4. This is a thorough, highly recommendable research that enlightens the rapport between Seville’s urban development and and its topographic context over time. As to its conclusions, they barely affect those reached herein, partly due to each of them dealing different scales.

• 5 VALENCIA RODRÍGUEZ, Rafael. El Espacio Urbano de la Sevilla Árabe. En AAVV, Premios Ciudad de Sevil (...)

11The departure point for the study of Islamic Seville’s evolution is the remaining of the wall inherited from Roman Hispalis imperial period. This has been a very much debated issue, due to the laconism and ambiguity of Muslim sources to its regard. Nevertheless, Valencia5 does not find evidence of the existence of intermediate fortifications between the Imperial Roman and the Almoravid-Almohad walls.

• 6 Ibid.

12There is, however, proof of punctual modifications on the Roman wall during the city´s Islamic period. The first main one being a restoration undertaken by initiative of Abd al-Rahman II as an answer to the Viking’s raids in the first half of the 9^th century, about which the sources do not give further details. The second mention in the sources refers to a programmed destruction of the walls by order of the Emir and soon Khalifa Abd al-Rahman an-Nasir, who had been compelled in the year 913 to reduce certain revolts in the city, in order to compromise the autonomy that it had been enjoying to the moment. Valencia aligns with A. Jiménez y J. de M. Carriazo6 in the belief that this demolishing was not to be fulfilled due to the protests of the citizens who, still in the memory of the Normand attacks, would feel exposed against new raids and, supposedly, the river’s floods.

• 7 COLLANTES DE TERÁN Y DELORME, (1977) op cit.

14It is also widely acknowledged that the city’s cardus maximus is slightly deviated eastwards. By examining a topographical plan like the one proposed by Collantes de Terán7, it is noticeable that the northern gate into Hispalis walls was placed in the highest spot between the basins of the river Tagarete and the ancient “urban” branch of the Guadalquivir, which ran through today’s Alameda de Hércules, Amor de Dios and Sierpes streets, to join back the main riverbed in the Arenal. The current situation of this gate is in the whereabouts of Santa Catalina’s church and its corresponding roman road would go by nowadays San Luis Street.

• 8 GESTOSO MOROTE, David et al. Informe preliminar de la excavación arqueológica preventiva en calle D (...)
• 9 RUIZ LÓPEZ, Urbano; GARCÍA MORALES, Irene. Excavación arqueológica preventiva en la c/ Joaquín Cost (...)

15Regarding roman times, the excavations undergone in the study area have detected disperse structures which have been associated to agricultural or maybe port uses. Specificly, remains of a farm of imperial period have been found in Divina Enfermera, 148; and of a late roman sidewall of temporary farm or dock use in Joaquín Costa, 329.

• 10 VALENCIA RODRÍGUEZ, R. (1988) Op. cit.
• 11 IBN HAYYAN, “Al muqtabas, V”. ed. CHALMETA, P; CORRIENTE, F. EN TAHIRI, Ahmed. Estructura Urbana de (...)

16Coming back to the discussion about the roman wall’s modifications between the 9^th and 10^th centuries, it seems that what really was carried out in 913 was a mere permeating operation, this is, aditional gates or wickets were opened10. Nevertheless there seems to be evidence of the prior -during the rule of Abd al-Rahman II on whose initiative Seville’s walls were restaured after the Normand attack in 844- opening of a second gate north of the wall11 identified as Bab al-Hadid.

• 12 GUERRERO LOVILLO, José. Al-Qasr al-Mubarak, el Alcázar de la bendición: discurso de recepción leído (...)
• 13 VALENCIA RODRÍGUEZ, Rafael. La Sevilla del Emirato Omeya. En AAVV, Aula Hernán Ruíz. De Hispalis a (...)

17The number and distribution of the resulting gates ensemble is object of discussion, being the two main hypotheses those of Valencia and the previous formulation made by José Guerrero Lovillo in his study about the abbadi al-Mubarak palace 12. They are both inclined to spot the new redundant gates on the west and east walls, though in his last statement to this respect, Valencia considers the possibility of having them opened on the north side13.

• 14 CASAL GARCÍA, Mª Teresa. Los cementerios musulmanes de "Qurtuba". Córdoba: Universidad de Córdoba, (...)

18The option of having a new gate opened in this part of the city as well gets much more strengthened by the recent finding of a 10^th century cemetery during an archaeological excavation in Divina Enfermera Street. This new piece of information strongly suggests the existence of a gate in its vicinity, as medieval Muslim graveyards would usually be located immediately out of the city boundaries and relate to gates and access roads14.

• 15 Ibid.

19There are other typical characteristics of hispano-muslim cemeteries of the time that can be reckoned, as the fact of being located on a hillside and nearby a water course 15. Against intuition, Cordoba’s al-Rabad cemetery shows that no inconvenient was found in practicing burials on terrains liable to be flooded. The graveyards were irregular in shape, they were not enclosed or fenced and could grow both by extension or accumulation. Hence they could contain several levels of burials, as it is the case of the one excavated in Divina Enfermera Street, particularly in their most appreciated locations, next to gates or sanctuaries.

• 16 IBN 'ABDÚN, Muhammad b. Ahmad. Sevilla a comienzos del siglo XII: el Tratado de Ibn 'Abdún. Emilio (...)
• 17 GUERRERO LOVILLO, J. Op. cit.
• 18 TABALES RODRÍGUEZ, Miguel Ángel. Excavación Arqueológica en el Jardín del Inglés del Real Alcázar d (...)
• 19 GAMARRA SALAS, Francisca E. Informe de la excavación arqueológica preventiva en la calle Santa Rufi (...)

20Ibn ‘Abdun’s treaty16, dated late 11^th or early 12^th century refers to the city’s cemeteries conditions. It mentions two of them, namely “the city’s” and “the potters’ graveyard” in the whereabouts of Bab Hamida gate, near the current postion of the cathedral. The theory placing there this gate, opposed to that of Guerrero Lovillo17, stands strengthened after the finding of abandoned kilns underneath the Alcazar18. Regardless of the exact location of the cemeteries mentioned by Ibn ‘Abdun, it is of interest how they are described. They appear -at the moment preceding the definitive enhancement of the city walls- crowded to the point of practicing new burials on top of former ones. This is confirmed by the recent excavations, which have documented superimposed entombments or even bodies that were moved when depositing others19.

• 20 CAMPOS CARRASCO, Juan Manuel et al. Excavación en c/ Joaquín Costa, nº8, Sevilla. AAA/1986. Tomo II (...)
• 21 The excavation was rather fruitless on the other hand, as it went only 1,60 m. deep before getting (...)

22The main supporter of the gate’s being in S. Martín square is archaeologist Campos Carrasco20 whose attention was driven towards the characteristic fan-shape distribution of streets Joaquín Costa, Juan Pérez de Montalbán, Correduría y Conde de Torrejón, during an emergency archaeological excavation in Joaquín Costa Street21. Due to the confluence of these streets’ south ends on the Europa Square, Campos ventured the theory of them being evidence of an ancient road network leading to a gate situated in a narrowing in Divina Enfermera Street, next to S. Martín church. Campos proposes that it be Bab al-Hadid, a gate that the sources associate to a marketplace which Campos identifies with the ensemble of long narrow blocks between the mentioned streets. For Campos, who eludes the specific position of the gate, this suburb and its souk took shape due to an access to the Abbadi palace of al-Mukarram, which would be built between the wall and the current Conde de Torrejón street.

• 22 GUERRERO LOVILLO, J. Op. cit.

24In that sense Guerrero Lovillo’s theory22, understanding that the Al-Mukarram palace was inside the walls in that same corner, seems more appropriate as it would occupy a topographically advantageous position and concentrate military power in a strategically relevant spot. This is an interesting possibility not only in order to locate this new gate on the wall, but also for the implications it might have regarding the urban development of our study area, as we will see later on.

• 23 VALENCIA RODRÍGUEZ, R. Op. cit.

28We should then decide on the option of, had there been a second gate in the north pane of the wall, having had it open in the S. Juan de la Palma square. This is a compatible hypothesis with the scarce sources which refer to the gate Bab el-Hadid. According to Ibn Hayyan23, the hayib Badr entered the city through this gate, in his way from Cordoba along a road that passed by Tocina, this is, following the route that, along the riverside, passed by Alcalá del Río.

• 24 BOSCH VILÁ, Jacinto. La Sevilla islámica: 712-1248. 2ª Ed. Sevilla: Universidad de Sevilla, Servici (...)
• 25 IBN HAYYAN EN TAHIRI, A. 2009 op. cit. p.18
• 26 Capital city of the Rif’s Emirate of Nakour, located in the whereabouts of Alhoceima Bay, Morocco, (...)

29This chronicle relates the bargaining of the city’s submission to the authority of Cordoba, and Bab al-Hadid has an important role in it. Not only can a tentative location be assumed hence, but also the fact of the gate’s existence prior to the 10^th century. As noted before, we know that the wall’s restoration works were undertaken during Abd al-Rahman II’s rule. This is not an isolated intervention if we considered that already in 830 the Khalifa had promoted the construction of the Grand Mosque of Ibn Adabbas, located where now El Salvador church is, and a dockyard24. Moreover, we know25 that he summoned Muhammad Ben Said Ben Abi Sulayman, from the city of Nakour26, to assist him in urban planning tasks; this may indicate that these projects on the city were not disconnected initiatives, but rather followed a common strategy.

• 27 En VALENCIA RODRÍGUEZ, R. (1988)

30In a different occasion27 it is told how rowboats passed between the Ibn Adabbas mosque and Bab al-Hadid’s market during the year 1200’s flood. Beyonf the anecdote, this narration would attempt, in my opinion, to give information on the magnitude of the flood which, in other words, would have inundated all the lower parts of city.

• 28 CHALMETA (1973) EN VALOR PIECHOTTA, Magdalena. Sevilla Almohade. Málaga: Sarriá, 2008. p. 160.

31As for the marketplace referred by the sources, it is to consider that a gate in S. Juan de la Palma would connect directly with the commercial axis of Regina Street, Encarnación and Siete Revueltas, leading to El Salvador. During a first expansion moment, should this be a marketplace linked to local commerce, its suppliers coming from the orchards north of the city and the Guadalquivir’s lowlands, sailing the river or by the riverside. We could even suppose that the weekly outer market, typical of Hispanic-muslim settlement networks28 had been removed from the city’s most prosperous commerce axes, to be accommodated out of this gate, having survived until nowadays in the shape of El Jueves.

• 29 VALENCIA RODRÍGUEZ, R. (1988) op. cit.

32It is not within the scope of this study to locate the rest of the imperial-islamic walls’ gates, though it is acknowledged that suggesting a new location for Bab al-Hadid affects the general distribution of the defensive ensemble. Hardly for Guerrero Lovillo’s hypothesis, but indeed for the one formulated by Valencia in 1986. The first one proposes a gate adjacent to the Al-Mukarram palace`s complex, though no name is specified. It seems logical that the Alcazar had a nearby exit, so the suggested gate by the author could be substituted by the one proposed in this article, accomplishing the same function. Or maybe it is a best option, given that a gate in the west wall would need to overcome an important height difference, as well as the remains of the river’s branch. In Valencia’s first hypothesis29 Bab al-Hadid is located westwards, in the vicinity of today’s Villasís square, and option which, provided that the theories herein be proved, should be reconsidered.

• 30 Op. cit.
• 31 RUÍZ LÓPEZ, U; GARCÍA MORALES, I. (2002) op. cit.

35Other conditions are needed for the understanding of the new suburb’s genesis. To the west the river’s urban branch, even if it is in a recession process, still exercises a strong pressure in the shape of frequent floods. In fact Borja Barrera and Barral Muñoz, after a geological study included in an archaeological excavation in Peral Street, assert that “during the centuries 6^th to 11^th an episode of peaking fluvial dynamics could be identified, followed by general xericity conditions or a recess of the flood pattern in all of the alluvial plains”30. This is a point that could be confirmed on the reports depicting recent archaeological excavations in the area. Whenever it was possible to date the strata, large height differences of up to two or three meters were found between roman and early medieval levels31, but just a few tens centimeters between these latter ones and later medieval layers.

• 32 Op. cit.

36On the other hand, from the 10^th century onwards the city experiments a demographic growth, still unmet by a new walling. This is the condition to the appearance of suburbs. Moreover, we know by Ibn ‘Abdun’s treaty’s32 pointing out the need to expel the potters out of the walls, that industrial activities, or those otherwise molest or polluting were moved outwards and replaced by more noble ones.

• 33 Both in Collantes de Terán map and EMASESA’s tacheometric data, these contour lines are altered by (...)

38As noticed by Campos Carrasco, the morphology of the sector enclosed by the streets Feria, Castelar, Pasaje Morgado, Perís Mencheta and the Alameda de Hércules is very characteristic. If attention is paid to the detailed topography in any of the historical maps (Guerrero Lovillo’s sketch is particularly expressive but also Collantes de Terán’s or any in the historical plannimetry of Seville is of use), these streets will appear to occupy a small depression on the left bank of the river’s urban branch, and to approximately draw the contour lines of this small topographic accident: curves that get more accentuated as we move from the Alameda towards Feria Street33.

• 34 This approach was already used by Alfonso del Pozo y Barajas in his morphological study of Seville’ (...)

39It does not seem to have there been a steep enough slope for a terraced development, therefore in my opinion these shapes are the result of a progressive occupation of the riverside space, as the flood line retracted34. We happen to gather here the conditions which in my opinion should intervene so the resulting urban tissue whould show this process so clearly:

• 35 CORTÉS JOSÉ, Joaquín, et al. Diccionario histórico de las calles de Sevilla. Sevilla: Consejería de (...)

40Regarding the last three points, Joaquín Costa Street, last to having taken shape according to this theory, presents a very expressive instance. There are records abut this street since the 15^th century, when it used to be named Cañaverería35. This was due to hosting the cane artisans guild, who would gather reed in the wetlands next to the city, probably around the Alameda de Hércules as well (a lagoon at the moment), for storage in their backyards. In the specific case of this street it is easily appreciated (particularly in Olavide’s 1776 map) how irregular the back walls of these lots are. Undoubtedly, the terrain’s evolution was stopped after the construction of the almoravid-almohad wall and the decrease of the river’s activity. These back walls would then freeze in such a state until the filling of the lagoon in 1574 and the later arrangement of an aligned façade towards the Alameda.

• 36 Ibid.
• 37 EN ESCARTÍN GONZÁLEZ, Eduardo. Economía y Sociedad en la Sevilla Almorávide. Sevilla: Secretariado (...)

41Remaining toponyms still witness the existence to similar activities36. Current Molino Street is named Molino del Yeso (plaster mill) in Olavide’s map; Cruz de la Tinaja street was named Tinaja or Tinajas (earthenware jar) up to mid 20^th century; there is notice about a charcoal scale in the current Peris Mencheta Street –Ibn ‘Abdun also talked about weighting charcoal by the river37-; in the 12^th century Quintana Street was known as Albardería, referring to the manufacture of pikes and Correduría Street reminds of leatherware guilds.

• 38 Op. cit.

43According to Borja and Barral’s geo-archaeological study38 and the reviewed archaelogiccal excavations, it is easy to establish the ending of this morphogenetic process at some moment in the 12^th century. The beginning would correspond to the opening of the Bab al-Hadid gate and consolidation of the riverside road between it and the later street Bibarragel. I have fixed this moment at the middle of the 9^th century seeking for coherence with the corresponding sources.