Zhang, 1999; Li et al. 2002; Zhang etal. 2005. Y en todos ellos se parte dela premisa de que debe existir un buenregistro entre la PAN y la MS, enten-diendo por buen registro la diferenciaposicional entre PAN y MS con valo-res inferiores al subpixel.

Amolins et al. (2007) clasifican losmtodos de fusin en clsicos, basa-dos en wavelet e hbridos, a los que ha-bra que aadir los multidireccionalesque recogemos en este trabajo (Lillo,2004) por su capacidad para detectarla anisotropa de la imagen. Los m-todos clsicos son capaces de detectarmuy bien la geometra de la imagenPAN e incorporarla a la MS pero tie-nen el inconveniente que durante elproceso de fusin introducen distor-siones del color. La distorsin del co-lor implica la asignacin de colores, azonas de la imagen fusionada, que noexistan en la MS original. Esta dis-torsin se acenta cuando la longitudde onda de las bandas MS no est cu-bierta por la longitud de onda de labanda PAN, como es el caso de la ban-

IntroduccinLas imgenes de teledeteccin (pro-

cedentes de satlites o areas) son usa-das a diario por cartgrafos, planifi-cadores del territorio y en general portodas aquellas disciplinas englobadasbajo la denominacin de Ciencias dela Tierra. Especial inters para el An-lisis Urbano tienen aquellas imgenesde la ciudad procedentes de satlitescon alta resolucin espacial como esel caso de las que se utilizan en nues-tro estudio (satlite QuickBird). Laimagen PAN tiene un tamao de pxelsobre el terreno de 0,7 m, mientras queel tamao de la MS es de 2,4 m. Porotra parte la resolucin espectral deQucikBird es de 4 bandas pertene-cientes a las longitudes de onda del in-frarrojo, rojo, verde y azul. En este tra-bajo slo se han empleado las bandasdel espectro visible: rojo, verde y azul(RGB). Por su parte la PAN es una ni-ca imagen pero que registra el anchode banda correspondiente al espectrovisible. Existe una gran cantidad detrabajos que estudian la fusin de im-genes de satlite p.e.: Nez, 1999;

182 LA FUSIN DE IMGENES DE TELEDETECCIN: UNA AYUDAPARA LA INTERPRETACIN DE ENTORNOS URBANOSJuan F. Reinoso Gordo, Carlos Len Robles

Palabras clave: Pancromtica, multiespectral, fusin, IHS, multirresolucin, wavelet, multidireccional, QuickBird.

La fusin de imgenes que interesaen la interpretacin de entornos urba-nos es aquella en la que se une unaimagen pancromtica (PAN) y otramultiespectral (MS). La PAN aportauna gran resolucin espacial pero s-lo en color gris mientras que la MSaporta el color producido por variasbandas (normalmente las tres bandasfundamentales: rojo, verde y azul). Elresultado es una imagen fusionadacon la mxima resolucin espacial yespectral. En este trabajo exponemosun ejemplo de los mtodos tradicio-nales como es la fusin mediante latransformacin IHS con indicacin desus problemas de distorsin del color,a continuacin se ejecuta una fusinmediante tcnicas de multi-resolu-cin basada en wavelets que solucio-na los problemas de distorsin. Final-mente se tiene en cuenta uno de losalgoritmos recientes que adems tie-ne en consideracin la anisotropa dela imagen. Para ello hace un anlisismultidireccional de las imgenes afusionar. La calidad de la fusin debeser analizada desde un punto de vis-ta cuantitativo, para ello se han em-pleado aquellas medidas que, en lasreferencias actuales, han mostradomayor utilidad. Un examen visual dela fusin tambin es necesario paraestablecer conclusiones sobre ca-da uno de lo mtodos. En nuestroestudio hemos empleado imgenesprocedentes del satlite QuickBirdpor ser uno de los que proporcionamayor resolucin espacial y portanto una mejor identificacin delos objetos del entorno urbano.

183da infrarroja en el caso del satliteQuickBird. Una solucin a la distor-sin del color se logr con el mtodomultirresolucin basado en wavelet(Nez, 1999), el cual era capaz de ex-traer de la PAN los detalles de la ge-ometra e incorporar slo dichos de-talles con lo que se conservaba en granmedida el color original de la MS. Latcnica de descomposicin wavelets enprincipio estaba basada en el algorit-mo de Mallat (1989), pero producaproblemas de anillo en torno a los ob-jetos singulares de la escena como con-secuencia del efecto Gibbs. Este efec-to de anillo se debe a que el algoritmode Mallat es diezmado: en cada eta-pa reduce el tamao de la imagen a 1/4de la imagen en la etapa anterior(Yocky, 1996). El efecto anillo se eli-mina mediante el uso de algoritmosno diezmado como es el caso del al-goritmo trous (Shensa, 1992), quesignifica con agujeros. Entre los m-todos multidireccionales ms recientesse encuentran algunas extensiones delas wavelets como son las tcnicas cur-velets (Cndes et al. 2000), contour-lets (Cunha et al. 2006) y otra des-arrollada por Lillo y Gonzalo (2004).

La evaluacin de la calidad de la fu-sin ha sido un tema ampliamente es-tudiado desde los inicios hasta la ac-tualidad. Se han desarrollado medidasque evalan tanto la calidad espacialcomo espectral de la imagen fusiona-da (Zhou, 1998; Wald, 2000; Wang,200, 2004). Estas medidas que cuan-tifican la calidad se ven complemen-tadas, generalmente, por un examenvisual que conduce a la emisin de unjuicio sobre los resultados.

A continuacin se presentan los m-todos de fusin (IHS, wavelet y mul-

tidireccional), las medidas empleadasy una imagen QuickBird sobre la quese aplican los mismos junto con unanlisis de resultados.

Fusin IHSLa fusin IHS se fundamenta en la

posibilidad de definir una imagen en co-lor mediante dos sistemas de coorde-nadas diferentes. El sistema ms habi-tual es el que almacena en tres matricesdiferentes cada una de las bandas de co-lor componentes rojo, verde y azul (sis-tema RGB); el otro sistema que nos in-teresa tambin almacena la imagen entres matrices que representan la inten-sidad, tono y saturacin (sistema IHS).En el sistema IHS la intensidad alma-cena la estructura espacial de la ima-gen, mientras que tono y saturacin al-macenan la estructura espectral. Aunquela imagen PAN se muestra en una es-cala de grises, tambin es posible des-componerla en el sistema IHS y comosu resolucin espacial es mayor, su com-ponente intensidad (I) conservar dicharesolucin. La transformacin consisteen sustituir la componente I de la MSpor la componente I de la PAN y la nue-

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

va imagen fusionada se vuelve a expre-sar en el sistema RGB. Para minimizarlas distorsiones del color, antes de rea-lizar la sustitucin la imagen PAN semodifica mediante la igualacin de suhistograma al histograma de la MS, ycon dicha PAN modificada se hace latransformacin IHS (Fig. 1). El paso delsistema RGB a IHS se expresa por lasecuaciones (1) y el paso de IHS a RGBse expresa por las ecuaciones (2).

1: Esquema de fusin IHS.

(1)

(2)

1. Para cada banda (R, G y B) de laoriginal MS se obtiene una nueva PAN(PANR, PANG, PANB) mediante la igua-lacin de su histograma con las refe-ridas R, G y B.

2. Se descompone mediante la trans-formacin wavelet cada banda de laMS original en detalles (RDetall., GDetall,BDetall) y aproximantes (RAprox., GAprox., BA-prox.) para un nivel de resolucin j, ennuestro caso j=2.

3. Se descompone mediante la trans-formacin wavelet cada PANR, PANG,PANB en detalles (PANR-Detall., PANG-Detall,PANB-Detall) y aproximantes (PANR-Aprox.,PANG-Aprox., PANB-Aprox.) para el mismo ni-vel de resolucin j que en el paso 2.

4. Por cada banda se crea una nue-va banda fusionada formada por losdetalles de la pancromtica (PANR-De-tall., PANG-Detall, PANB-Detall) y los aproxi-mantes de la MS (RAprox., GAprox., BAprox.).Lo cual da lugar a la fusionada colordefinitiva (ecuacin 5).

Fusin WaveletLa transformada wavelet puede des-

componer una seal (en nuestro casouna imagen) en dos partes, una co-rrespondiente a un aproximante (par-te grosera de la imagen) y otra corres-pondiente a los detalles (geometra delos objetos singulares). Matemtica-mente se puede expresar la imagen f(x)como en la ecuacin 3:

184

(3)

Donde cjk y djk son los coeficientesde aproximacin y detalle de las fun-ciones de escala (jk (x)) y wavelet (jk(x)) respectivamente. J indica el nivelde resolucin hasta el que se ha lle-gado en la descomposicin de la ima-gen. Las funciones de escala {jk (x)}y wavelet {jk (x)} forman bases orto-normales de subespacios Vj y Wj quea su vez son complementarios entre s.

Las formas de la wavelet madre y lade la funcin de escala, expresadas am-bas en forma didica, son las de lasecuaciones (4)

j,k (x)=2(2 j x-k); j,k;

j,k (x)=2(2 j x-k);(4)

RGBFusionada =[ RAprox. + PANDetall., GAprox. + PANDetall., BAprox. + PANDetall.]

(5)

(6)

j,k (x) tiene el efecto de un filtro depaso bajo mientras que j,k (x) tiene elefecto de un filtro de paso alto.

El algoritmo de fusion wavelet si-gue los siguientes pasos:

Vase el proceso grfico en la Fi-gura 2; para no aumentar excesiva-mente el tamao de la misma se haomitido la representacin de (PANR-Detall., PANG-Detall, PANB-Detall) y (PANR-Aprox.,PANG-Aprox., PANB-Aprox.) y slo se hapuesto a modo de ejemplo PANAprox. yPANDetall. que se han extrado de laPAN original.

El algoritmo empleado utiliza la fun-cin de escala de las B-Spline cbicasque se traduce en el filtro bidimensio-nal de la ecuacin 6. Dicho filtro se haaplicado con arreglo al algoritmo nodiezmado trous (Shensa, 1992).

185

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

MS color Original PAN Original PANAprox. j=2 PANDetall. j=2

R Original RAprox. j=2 RDetall. j=2 RAprox. + PANR-Detall.

G Original GAprox. j=2 GDetall. j=2 GAprox. + PANG-Detall.

B Original

RGBFusionada =[ RAprox. + PANDetall., GAprox. + PANDetall., BAprox. + PANDetall.]

BAprox. j=2 BDetall. j=2 BAprox. + PANB-Detall.

2. Secuencia de Fusin de imgenes mediante tcnica wavelet

F : valor medio de los pixeles en la banda fusionadaO: valor medio de los pixeles en la

banda original

Las medidas son las siguientes: Indice de Calidad Universal (ICU)(Wang y Bovik, 2002):

Medidas evaluadoras de lacalidad

Las medidas evaluarn la calidadtanto espectral como especial. Para ellose suele comparar la imagen fusiona-da, bien con la original en color, lo queda idea de lo bien que se ha conserva-do el aspecto espectral, bien con lapancromtica que dar informacinsobre la cantidad de detalles geom-tricos que se han incorporado a la ima-gen fusionada. Teniendo en cuenta esoscondicionantes se han seleccionado 7medidas de forma que cumplan conlos objetivos de la evaluacin.

La notacin empleada es la si-guiente:

i: iesima bandaOj : jesimo pixel en la iesima banda original. Tambin puede ser la PANFj : jesimo pixel en la iesima banda fusionadaNP: nmero de pixeles

Fusin Multiresolucin Multidireccional

Con este algoritmo se pretende de-tectar con ms precisin los detallesque no tengan orientaciones horizon-tales, verticales o diagonales, si no queocupen otras direcciones de la imagen(Lillo y Gonzalo, 2007). En este sen-tido aventaja al algoritmo del aparta-do anterior, que slo poda tener encuenta las tres direcciones menciona-das. En el aspecto de la ejecucin slose diferencia respecto del algoritmowavelet anterior en la forma de obte-ner los detalles, existirn tantos nive-les de resolucin como direcciones sedeseen analizar. Cada aproximante seconsigue mediante el filtro elipsoidalH(u,v) ecuaciones 7.

186

(7)

Donde es la direccin considera-da, a es la escala y b es la elongacin.

La Figura 3 presenta el esquema deobtencin de aproximantes y detallesdel algoritmo multidireccional. Unavez obtenidos detalles y aproximanteslas fases son las mismas que en el al-goritmo de la seccin 3.

3. Esquema de fusin multidireccional.

Coeficiente de Correlacin (CC):

187 ndice de Similaridad Estructural: mi-de la informacin estructural en la ima-gen fusionada (ISE) (Wang et al.,2004):

Para analizar y comparar los mtodosde fusin se han construido tres tablas: Tabla 1 (Medidas espectrales): se

calcula cada medida entre bandashomlogas (p.e. las bandas rojas)que pertenecen a la imagen fusio-nada y a la imagen MS. La medidase calcula para cada una de las tresbandas, lo que produce 3 medidaspor cada imagen fusionada, con esastres medidas se puede calcular la me-dia que ser el valor que se usar pa-ra comparar la bondad entre los m-todos. Cuanto mayor sea el valor dela medida (excepto en el caso de lamedida discrepancia) mejor se ha-br transmitido el color desde la MSoriginal hasta la imagen fusionada.

Tabla 2 (Medidas espaciales): lasmedidas se calculan de manera si-milar a las de la tabla 1 pero en es-te caso cada banda de la imagen fu-

sionada se compara siempre con laPAN. El valor de la medida indicael grado con el que los objetos dela PAN (con mucho mayor nivel dedetalle que en la MS) se han trans-ferido a la imagen fusionada. Cuan-to ms alto es el valor de la medidamayor resolucin espacial ha con-seguido la fusionada.

Tabla 3 (Medidas globales espacialesy espectrales): las medidas que apa-recen en esta tabla no realizan un an-lisis banda a banda, sino que pro-porcionan un valor global de lacalidad de la fusin. El aspecto espa-cial es medido por la medida ERGASespacial, mientras que el aspecto es-pectral es medido por las medidasSAM y ERGAS espectral. En esta ta-bla a diferencia de las anteriores,cuanto ms pequeo es el valor me-jor es la fusin.

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

M= nmero de ventanas locales de laimagen

Discrepancia (DC):

ERGAS (Wald, 2000):

donde IMG es PAN o Fh y l son las resoluciones espaciales dela PAN y MS respectivamente

Zhou index (Zhou et al., 1998)

donde PAN y F son las imgenes PANy la fusionada filtrada por el laplacia-no respectivamente

Mapa Angular Espectral (SAM)(Nencini et al. 2007):

Tabla 1: Medidas Espectrales

Medidas EspectralesIHS Wavelet Multidir

Calidad Universal Rojo 0,798 0,953 0,955Verde 0,811 0,952 0,954Azul 0,792 0,953 0,953

Media 0,801 0,953 0,954MSSIM Rojo 0,487 0,581 0,603

Verde 0,480 0,584 0,605Azul 0,472 0,563 0,578

Media 0,480 0,576 0,595Coeficiente de Correlacin Rojo 0,825 0,953 0,955

Verde 0,820 0,952 0,954Azul 0,826 0,953 0,953

Media 0,824 0,953 0,954Discrepancia Rojo 34,105 15,987 15,872

Verde 31,320 15,998 15,993Azul 34,535 16,542 16,836

Media 33,320 16,176 16,234

sin embargo tras la evaluacin visualde las imgenes de la Figura 4, en elmtodo IHS se aprecia distorsin delcolor, especialmente en los rboles yen tejados cuyo color es anaranjadoen la MS original. Tambin se obser-va una prdida de intensidad del co-lor rosa en la zona de la plaza. Losmtodos Wavelet y Multidireccionaltienen unos valores muy similares,tanto en su faceta espacial como mul-tiespectral, aunque de la observacinvisual, parece que las Wavelet repro-ducen con un poco ms de fidelidadlos colores. En cualquier caso, estosdos ltimos mtodos renen las cua-lidades que se le pide a la fusin deimgenes: integracin en una nicaimagen de la mayor resolucin espa-cial de la PAN y la mayor resolucinespectral de la MS.

7. ConclusionesLos mtodos de fusin de imge-

nes que se han presentado en este tra-bajo se muestran tiles en la integra-cin de informacin geogrfica,concretamente urbana, procedentesde imgenes de diferentes resolucio-nes PAN y MS. Particularmente per-miten la discriminacin de objetosque si se observasen en las imgenesPAN o MS separadamente no seraposible asignar correctamente a la ca-tegora que pertenecen. Esto implicauna ayuda considerable a aquellaspersonas que han de interpretar losentornos urbanos basndose en la ob-servacin de imgenes urbanas (ennuestro caso imgenes procedentesdel satlite QuickBird).

Ntese cmo todas las medidas nose aplican para las evaluaciones es-pectrales y espaciales, pues en algunosno tiene sentido hacer una interpre-tacin de la medida.

Anlisis y discusin de resultados

De la observacin de las tablas 2 y3 se tiene que los mejores valores es-paciales los obtiene el mtodo IHS;

188 Medidas EspacialesIHS Wavelet Multidir

Calidad Universal Rojo 0,961 0,908 0,917Verde 0,991 0,906 0,915Azul 0,965 0,864 0,876

Media 0,972 0,893 0,903MSSIM Rojo 0,940 0,849 0,860

Verde 0,973 0,849 0,860Azul 0,936 0,808 0,822

Media 0,950 0,836 0,847Coeficiente de Correlacin Rojo 0,985 0,912 0,920

Verde 0,996 0,908 0,917Azul 0,986 0,868 0,880

Media 0,989 0,896 0,906Zhou Rojo 0,995 0,989 0,986

Verde 0,998 0,984 0,982Azul 0,995 0,976 0,975

Media 0,996 0,983 0,981

Medidas Globales Espectrales y EspacialesIHS Wavelet Multidir

Espacial ERGAS 0,766 1,111 1,082Espectral ERGAS 1,328 0,931 0,925SAM 3,769 3,244 3,482

Tabla 3: Medidas Globales Espectrales y Espaciales.

Tabla 2: Medidas Espaciales.

189

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

4. Resultados de la Fusin. De izquierda a derecha yde arriba abajo: MS, PAN, Fusionada IHS, FusionadaWavelet y Fusionada Multidireccional.

Referencias

Y. ZHANG, G. HONG: An IHS and wavelet integrated ap-proach to improve pan-sharpening quality of natural co-lour IKONOS and QuickBird images, Inf. Fusion., 2005, 6(3), 225-234.

J. NEZ, X. OTAZU, O. FORS, A. PRADES, V. PAL, R. ARBIOL:Multiresolution-Based Image Fusion with Additive Wa-velet Decomposition, IEEE Trans. Geosci. Remote Sen-sing., 1999, 37 (3), 1204-1211.

Y. ZHANG: A new merging method and its spectral and spa-tial effects, Int. J. Remote Sens., 1999, 20 (10), 2003-2014.

S. LI, J. T. KWOK, Y. WANG: Using the discrete wavelet fra-me transform to merge Landsat TM and SPOT panchro-matic images, Inf. Fusion., 2002, 3(1), 17-23.

K. AMOLINS, Y. ZHANG, P. DARE: Wavelet based image fu-sion techniques - An introduction, review and compari-son, ISPRS-J. Photogramm. Remote Sens., 2007, 62(4),249-324.

M. LILLO, C. GONZALO: Multispectral images fusion by ajoint multidirectional and multiresolution representation,Int. J. Remote Sens., 2004, 26 (6), 1263-1268.

S. MALLAT: A theory for multiresolution signal decompo-sition: the wavelet representation, IEEE Trans. PatternAnal. Mach. Intell., 1989, 11(7), 674-693.

D. A. YOCKY: Artifacts in wavelet image merging. Opt.Eng. 1996, 35(7), 2094-2101.

M. J. SHENSA: The Discrete Wavelet Transform: Weddingthe trous and Mallat algorithms. IEEE Trans. SignalProc. 1992, 40 (10), 2464-2482.

E. J. CANDS, D. L. DONOHO: Curvelets- a surprisingly ef-fective nonadaptive representation for objects with ed-ges, in Curves and Surfaces, C. Rabut, A. Cohen, L. L.Schumaker, Eds.Vanderbilt University Press. 2000.

A. L. DA CUNHA, J. ZHOU, M. N. DO.: The NonsubsampledCounterlet Transform: Theory, Design, and Applications.IEEE Trans. Image Process., 2006, 15 (10), 3089-3101.

J. ZHOU, D. CIVCO, J. SILANDER: A Wavelet method to mer-ge Landsat TM and SPOT panchromatic data, Int. J. Re-mote Sens., 1998, 19 (4), 743-757.

Z. WANG, A. C. BOVIK: A Universal Image Quality Index.IEEE Signal Process. Lett. 2002, 9(3), 81-84

Z. WANG, A. C. BOVIK, H. R. SHEIKH, E. P. SIMONCELLI: Imagequality assessment: from error visibility to structural simi-larity, IEEE Trans. Image Process., 2004, 13 (4), 600-612.

Wald: Quality of high resolution synthesized images: Isthere a simple criterion? in Proc. Int. Conf. Fusion of earthData., 2000, Jan, 99-105.

F. NENCINI, A. GARZELLI, S. BARONTI, L. ALPARONE: 'Remotesensing image fusion using the curvelet transform', Inf.Fusion., 2007, 8 (2), 143-156.

scene-painter he was, he had no hesitation in some-times removing and adding elements of architec-ture or landscape in the faithful travel sketches, withthe aim of accentuating theatrically the effects ofthe composition, or contrasts of outdoor and indoorlight (evenings, night views ), sometimes recon-structing lost architecture15, skewing proportions,altering the scale of the human figures, or inventingelements that were not included in the sketches be-cause of a lack of time. In spite of all this, Robertswas always true to the essence of what he repre-sented. To use an old adage applied to the skilfulbullfighter, in his drawings Roberts sometimes de-ceives but without lies. It is therefore inaccurate to state of the views, withCalvo Serraller, that: in these representations ofSpanish cities the fantastic predominates over thereal 16, since exactly the opposite is the case, thereal predominates over the fantastic. This is of greatimportance when consideration is given to the doc-umentary value of the illustrations, the details ofwhich need to be analysed one by one with an un-prejudiced eye. Likewise open to question is the opin-ion of Galera Andreu on Roberts: objectivity, real-ism or accuracy are not values much taken accountof in this way of looking at and reflecting landscapeor scenery17. Among the numerous examples whichdemonstrate the opposite are lithographs of the Gi-ralda or the Tower of Comares (1837) (fig.6), whichdespite containing certain alterations, are among themost beautiful and faithful representations of thesemonuments prior to the advent of photography. Thuslabels or inappropriate generalizations should beavoided and greater rigour applied to the views tak-en individually and compared with other drawings ofthe 19th century (Laborde, Ford...) or where possi-ble, with photographs so as to be able to make pre-cise judgements in each instance on their veracityor otherwise.In any case if one had to sum up Roberts and hiswork, it can be affirmed that he was a great painterof landscapes and architecture, a graphic architectwith expressive points of view and sophisticated re-sources in his scenes, a true landscape artist, andnot a Romantic, since such a label implies a par-tial vision of a rich body of work. The term Roman-tic has proved productive in literary research butis questionable when applied to the graphic arts,according to Honour18. Roberts never engaged in lit-erature, unlike other Romantic travellers whowrote with a teeming imagination, and sometimesgave picturesque or antiquated versions of reality.An entry of 16 March 1833 in the diary of the BritishConsul in Malaga, William Mark, personal friendof Roberts, who accompanied him around the city,states clearly: Mr Roberts has come to Spain tomake a faithful study and to distance himself in his

subjects are not only good but of a very novel char-acter. I begin to doubt whether I shall be able to painthalf of them, they already amount to over 100, and Ihave yet six towns to visit, from which I expect manyfine subjects. In another letter two months later,while passing through Gibraltar, Roberts remarks:many of the drawings are very original, I have nofewer than two hundred and six finished drawings,the greater part coloured, in addition to others in mysmall sketch-book8. It has been estimated that in An-dalusia alone Roberts made as many as 250 views9.However, he would not paint oils in Spain, with theexception of two in Seville, on account of the obvi-ous difficulties involved in transportation to England.Given that his sketch-books were pocket-sized (10by 17 cm.), Roberts ability to use them to create com-plex and detailed scenes is remarkable. Accordingto Katharine Sim, what is astonishing in thesesketches, the majority in pencil or wash on lightly tint-ed paper is not only the simplicity and delicacy of theline, but also the surprising accuracy of subtly ren-dered details10. According to Ballantine, who was a personal friendof the painter, He seemed to have the power of pho-tographing subjects on his eye, for I have again andagain been with him while he was sketching veryelaborate structures or very extensive views, and hetook in a large mass at one glance, not requiring tolook again at that portion until he had it completed inhis sketch. Other artists caught small bits at a time,and required to be renewing their glances continu-ally. Roberts, by this extraordinary faculty, either nat-ural or acquired, got over more than double their workwith half their labour11. The rapidity required in thescene-painting of his youth would explain howRoberts came to have such facility in capturing andrecording his images. Following his extended visit to Spain Roberts pro-duced numerous views of the same scenes, repro-ducing again and again those most in demand fromhis clients, with variations which, like a propergraphic architect, he introduced with great preci-sion, with the aim of satisfying the requests he re-ceived. Arguably the distortions or manipulations ofthe views were the result not of a taste for the fan-tastic or an allegedly Romantic character, whichhas been the object of so much speculation12, but ofthe requirements of those who paid him for worksthat were unique and exclusive, elaborated on thebasis of small sketches, sometimes long after thejourney itself (1833), as was the case with a view ofCarmona, dated 1853, which Roberts gave as a pres-ent to his dentist13. According to Sim, the painters biographer, what in-terested him was always to take advantage of his ex-traordinary ability as draughtsman, rapid and faith-ful to reality14. Nevertheless, like the excellent

1. Education and the background to the tour of SpainDavid Roberts (1796-1864) was born into a poor fam-ily in Stockbridge, near Edinburgh: his father was ashoemaker and his mother a washerwoman. Robertshad to struggle to make his way in his profession.At the age of twelve he entered as apprentice in theworkshop of painter and decorator of Scottish coun-try houses, and after struggling for seven years hefound employment as a painter of stage scenery in atravelling theatre, which allowed him to travel andpaint scenes on a large scale and to draw the placeshe visited. According to the author himself, it waswhile drawing York Cathedral that he discovered histrue vocation as a painter-traveller5.From 1816 until around 1830 Roberts devoted himselfto scene-painting and worked in the most importanttheatres in Scotland and England, executing numer-ous graphic compositions with great inventiveness,immediacy and expressive intensity, all of which al-lowed him to mature as an artist. As his financial cir-cumstances improved, he was able to devote himselfto painting. In 1824 he became a member of the new-ly formed Society of British Artists6, where he exhib-ited many paintings relating to his travels. After ex-hibiting West Front of Notre Dame, Rouen at the RoyalAcademy, painted for Lord Northwick, one of his firstand most generous patrons, Roberts received a num-ber of very attractive commissions and in 1831 theSociety of British Artists elected him as their presi-dent.Among the reasons that inspired his journey to Spainwere Roberts own personal circumstances. He mar-ried in 1820 and his only daughter was born a yearlater. The failure of his marriage and subsequent sep-aration led him to seek a change in both his person-al life and artistic direction7. Around 1830 Roberts wasin an ideal condition, physically and financially, aswell as in terms of age and travelling experience, toundertake an adventure in Spain. Some of his friendsand contemporary British artists such as David Wilkie,John Frederick Lewis, or other figures, like RichardFord had done the same. The opportunity to stand be-fore works by masters such as Velzquez, Murillo orZurbarn exercised an obvious appeal. At the timeSpain was in vogue and there was considerable pub-lic demand in England for all things from Spain, thenlittle known or considered exotic. This convincedRoberts that the substantial effort and expense in-volved in the journey would be amply rewarded bythe sale of his Spanish work on his return, as indeedwas the case.

2. A prolific painter of landscapes and architectureRoberts embarked on his journey with the intentionof making the fullest possible use of his time, and hadpencil or paintbrush always to hand. After scarcelysix weeks in Spain, in a letter written from Crdoba,Roberts could write: My portfolio is getting rich, the

5 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 21 etseq., 2002.6 / QUESADA, L.: Pintores espaoles y extranjeros enAndaluca, p. 86 and ff., Sevilla 1996.7 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 28 etseq., 2002.

14 / SIM, K.: David Roberts R. A., 1796-1864. A Biography, p.80, 1984.15 / GAMIZ GORDO, A. / ANGUIS CLIMENT, D.: Imgenescordobesas: los Molinos y la Albolafia, El Legado Andalusn 22, p.80-89, Granada 2005.16 / CALVO SERRALLER, F.: La imagen romntica de Espaa.Arte y Arquitectura del siglo XIX, p. 22 y ss., 95-96, 1995.17 / GALERA ANDREU, P.: La imagen romntica de laAlhambra, p. 134-37, 1992.18 / In the visual arts there is no such thing as a Romanticstyle if by this we mean a common language of visual formsand means of expression comparable to the BaroqueRococo. There is not a single work of art that embodies the

8 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 56 y218, 2002.9 / QUESADA, L.: Pintores espaoles y extranjeros enAndaluca, p. 87, 1996.10 / SIM, K.: David Roberts R. A., 1796-1864. A Biography, p.68, 1984.11 / BALLANTINE, J.: The life of David Roberts R. A.,p.231, 1866.12 / GALERA ANDREU, P.: La imagen romntica de laAlhambra, El Viso 1992.13 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 108(cat. 85), 1986.

204

205

objectives and ideals of the Romantics .... HONOUR, H.: Elromanticismo, p. 15-16, 1981; QUESADA, L.: Pintoresespaoles y extranjeros en Andaluca, p. 72, 1996.19 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 185-186,2002.20 / ROSCOE, T. (text) / ROBERTS, D. (art.): The tourist inSpain, R. Jennings, 4 vols. London 1835-38. Among thenumerous editions there was a French edition of thevolume dedicated to Granada entitled LEspagne. Royaumede Granada, 1835.21 / A copy exists in the Barbican Art Gallery. GIMNEZCRUZ, A.: La Espaa pintorescap. 369-70, 2002.22 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 48,

57, 107-108 (cat. 81, 82, 83, 84), 1986.23 / MESONERO ROMANOS, R.: Manual de Madrid, p. 13-26(2 ed.), 1833.24 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 83, 2002.25 / Letter sent from Crdoba 30 January 1833. GIMNEZCRUZ, A.: La Espaa pintorescap. 102, 2002.26 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 60,61 y 108 (cat. 86, 87, 88, 89), 1986. CATLOGO: ArtistasRomnticos Britnicos en Andaluca, 2005.27 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 141-143,2002.28 / FORD, R.: Gatherings from Spain, p. 293, Madrid, 1974.29 / Towneley Hall Art Gallery and Museum, Burnley. Pudo

basarse en dibujo de R. Ford conservado por la familiaFord.30 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 173-177,2002.31 / It appears to be based on a drawing by R. Ford in thepossession of the Ford family. The watercolour isreproduced in SIM, K.: David Roberts R. A., 1796-1864. ABiography, 1984. GUITERMAN, H. / LLEWELLYN, B.: DavidRoberts, p. 109 y 110 (cat. 96 y 97), 1986.32 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 215,2002.33 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 109(cat. 90), 1986.

(1835) in the Fitzwilliam Museum, another in the Na-tional Museum of Wales, and an oil (1834) in Wake-field Art Gallery. The other five originals are devotedto the Mosque: the portico of the Courtyard (1833) ba-sis of the above-mentioned lithograph, and thechapel of Mahomed (1833) together with anothersimilar water-colour in private collections; The Sanc-tuary of the Koran (1849) in Leicester Galleries; andan oil in the Prado (fig.5)26. From Crdoba Roberts continued his journey anddrew views for engravings (1835) en Luque, and Al-cal la Real, arriving in Granada in February 1833.He remained there two weeks. Awed by the city, itsposition and the Alhambra, he remarks in a letter tohis sister that there are so many beautiful objectsthat I am at a loss which to take first. My breakfastis generally over by eight oclock, when I go out todraw. I dine at two, and out to work again. AlthoughI have nobody to speak to, I never feel the time long;indeed, if I feel inclined to grumble at all, it is at theshortness of the days.The architecture is so pecu-liar and elaborate that it would take months to doit justice []. And now I am going to smoke a ci-gar and go to bed, to dream of Moors and Christians,tournaments and battles, painting and architecture.In the morning I go to the Court of the Lions []andI mean to finish a drawing of it which has alreadytaken me all this day27.Granada was the Spanish city that inspired the great-est number of engravings (1835), a total of fifteen: ageneral view of the Alhambra, Granada from thebanks of the Xenil, the Palace of the Generalife, theVermillion Towers, Descent into the Plain and Viewof Granada, Tower of Comares, the Gate of Justice,the Court of the Alberca, Remains of the Old Bridgeacross the Darro, the Casa del Carbn, the Tower ofthe Seven Vaults, the Hall of the Abencerrages, Moor-ish Gateway leading to the Viva Rambla, Court of theLions and the Hall of Judgement. There are in addi-tion five lithographs (1837): Entrance to the Ferdinandand Isabel (plate I), the Casa del Carbn (pl. IX), theTower of Comares (pl. X) (fig.6), Old Buildings on theBanks of the Darro (pl. XI) and the Tomb of theCatholic Monarchs (pl. XVI). There is a rare engrav-ing of the Courtyard of the Lions (fig.7) and other orig-inal views, the subject of a forthcoming monographby the present author.Roberts was forced to leave Granada when he wasprevented from continuing a panoramic view of thecity and the surrounding country from the Alhambra.This was the only setback Roberts encountered onaccount of the intransigence of the military authori-ties, which was frequent in the period, as the fol-lowing graphic comments of Richard Ford make clear:Nothing throughout the length and breadth of theland creates greater suspicion or jealousy than astrangers making drawings, or writing down notes

terior of the Catedral (1836) basis of the engraving of1837; and in a private collection there is another out-door view in pencil (1836), which gave rise to a lith-ograph in 1837 (pl. VII). There is an oil of the Staircasein the North Transept (1835) in the Tate Gallery,and awater-colour (1836) in the Whitworh Art Gallery22, onwhich two variant engravings are based. A drawingof the Chapel of the Condestable (1845) can be foundin The Courtauld Institute of Art, and an oil-paintingon the same subject in the Prado in Madrid.Roberts arrived in Madrid on 16 December 1832 andspent the Christmas period there. According toMesonero Romanos23, the city had a certain small-town charm, in comparison with European capitalssuch as London, Paris or Rome. In a letter to his sis-ter the town consists almost entirely of churchesand convents and I have made a number of sketch-es24. Perhaps he missed a large cathedral like thatof Burgos. Roberts enjoyed the great Spanish mas-ters in the Prado (Velzquez, Jusepe de Ribera),met painters of the period such as Federico Madra-zo or Vicente Lpez, and was critical of the stateof landscape painting in Spain at the time: its aninsult to give it that name25. His three-week staygave rise to six engravings (1837): Entrance toMadrid by the Gate of Fuencarral, Street of SanBernardo, Street of Alcal, the Royal Palace, theHigh Altar, Church of San Isidro, and the Fountainon the Prado. In addition he made two lithographs(1837): the Gate of Alcal (pl. XVIII) and Bridge ofToledo on the Manzanares (pl. XX). Two originaldrawings are known: one of the Bridge of Toledoin The Royal Collection, and a water-colour of theFountain on the Prado, basis of the above-mentionedengraving in the Denver Art Museum.Roberts left Madrid on 9 January 1833 and arrived inCrdoba on 12 or 13 January. He remained there threeweeks and fruit of his labour were four engravings:(1836): Cordova: Looking down the Guadalquivir, In-terior of the Great Mosque, The Alcazar or Prisonof the Inquisition, and The Tower of the Church ofSt Nicholas; also five lithographs (1837): a generalview of the bridge and the river with the Mosque inthe background (pl. XXIV), a watermill and the bridgeover the Guadalquivir (pl. XXII), the entrance gateto the city (pl. XII), the portico of the Courtyard of theMosque (pl. XXI), and the chapel of the Mosque (pl.XIII). Later he recorded another general view of thebridge and the Mosque. A total of twelve originals ofCrdoba by Roberts are known, of which seven areoutdoor views: a water-colour of the Alcazar in Dud-ley Museum; another of the Alczar, basis of theabove-mentioned engraving, in the Whitworth ArtGallery; a drawing of the Albolafa belonging to a pri-vate collection in Granada; an oil of the Tower of Cala-horra in another private collection; three more ver-sions of the Church of St Nicholas, a water-colour

work from the merely picturesque, and is in this re-gard unsurpassed19.

3. The itinerary in Spain and the views executedIn what follows, an account is given of Roberts itin-erary in Spain and the views he executed, first thepublished illustrations followed by the known origi-nals. It should be mentioned that the whereabouts ofmany of the originals are unknown or they were re-worked for publication as the twenty-six lithographsof Picturesque sketches in Spain. Taken during theyears 1832-1833 (1837) and the approximately sev-enty steel engravings (smaller vignettes, omitted herefor reasons of space) included in the four volumes ofLandscape Annuals edited by R. Jennings with thetitle The tourist in Spain20 (1835, 1836, 1837 and 1838).For this project the finest engravers of the time wereemployed and use made of original drawings not bythe author himself, as we shall see.Regarding the route followed by Roberts on his jour-ney through Spain, many details can be gleaned fromhis letters as well as from a personal diary known asthe Record Book21, in which he noted down variousmatters, such as the paintings he was working on,buyers and sellers etc. The diary records the firststages of this trip in 1832, which was marked by dis-astrous epidemics: On October 18th, I left Londonfor Spain and after some days in Paris, continued onmy journey, passing through Orleans, Tours and Bor-deau, remaining in this last for some time. From thereI proceeded to Bayonne where I was detained inquarantine for twenty days. I was finally allowed toleave and continue my journey and enter Spain. Onarriving in Irun, the first Spanish town, I was in alazaretto for fourteen days on account of the cholera,then very prevalent in France. In the month of De-cember I crossed the Pyrenees and after passingthrough Vittoria, came to Burgos.There is little information concerning the initial stagesof Roberts journey. Volume III of The Tourist inSpain (1837) contains three engravings of theBasque Country: Fontarabia from St Jean de Luz, aView on the Bidassoa looking towards Irn, and theGreat Square of Vitoria, where Roberts spent a night.En route to Burgos, Roberts made a view of Mirandadel Ebro and another of the Pass of Pancorvo. A del-icate original drawing, reproduced here (fig.2), is pre-served in a private collection in Seville. It depicts anunidentified landscape resembling that of Pancor-vo and was no doubt executed with the intention ofproducing an engraving, although this was never ac-tually published.Roberts spent a week in Burgos and volume III in-cludes five engravings (1837): Entrance to the city ofBurgos, Ruins of the Convent of the Carmelites, WestFront of the Cathedral (fig.3), the Staircase in theNorth Transept and Tower of the Chapter House. TheBritish Museum contains a water-colour of the ex-

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

visited these places from Madrid, Gimnez Cruz41 be-lieves it is unlikely, since they are not mentioned inthe Record Book, and in the letter written from Cr-doba he expressed his intention to visit them on hisreturn journey, which he finally made by ship. Thissuggests that he did not actually go to these placeshimself. It should be pointed out that it was RichardFord who offered Roberts his own drawings, whichhe was able to use in Volume I (1835) for the engrav-ings of Loja and Ronda already mentioned42. In volume IV (1838) the following authors of thesketches for the engravings are cited43: Richard Ford;Colonel Harding, and another two military men, Smithand Edridge, of the Royal Engineers. According toKatharine Sim it is very likely that one of his moresensitive friends perhaps the resourceful RichardFord mentioned [to Roberts] these omissions in theAnnual corresponding to 1837 [vol. III] and suggest-ed that it would be more sensible to include the dueacknowledgements44. Three engravings were published of Segovia in vol-ume III (1836), City of Segovia, the Great RomanAqueduct, and the Alczar; and another view in vol-ume IV (1838). The following originals by Roberts areknown: a water-colour of the Alczar (1836) in Birm-ingham City Museum, a water-colour of the Aque-duct in the British Museum, and another on the samesubject in the Whitworth Art Gallery45. With regardto the El Escorial, an engraving was published (1836)and a lithograph (1837) (pl. XXIII). According to Sim46,the views of Segovia and El Escorial contain incon-gruities that lead one to believe he did not makethem directly himself. Regarding the engraving ofthe view of Toledo (1837), his preparatory drawingbelongs to the Ingram Family Collection, and seemsto have been inspired by a sketch by Sir EdmundHead47. There is a variant of this engraving (fig.4)with alterations in the foreground.Other engravings appear in volume IV (1838): Sala-manca, from above the River Zerguen, on an originalby Richard Ford48; a view of Plasencia also based onsketches by Ford49, although Roberts original, the ba-sis of the engraving in the British Museum, survives;the same occurs with the engraving of Santiago deCompostela (1838)50. As is indicated in the engravingsthemselves, on the basis of sketches made by thearmy officers already mentioned three views of Va-lencia were executed: Plaza Cathedral, the Gate ofthe Serranos, and Tower of Santa Catalina; togeth-er with the view of the Leaning Tower at Zaragoza.

5. The great success and popularity of the viewsDavid Roberts would never return to Spain, thoughon his return to England, at the height of his powersas a person and an artist, and until 1838, he devotedfive years to exploiting his travel sketches. Robertsreceived an astronomical sum for The Tourist inSpain (1835): for twenty drawings and the cover, with

painters and exercised a notable influence on thelandscape painter Jenaro Prez Villa-amil36. Sevillewas the subject of seven engravings (1836): Plaza Re-al and the Procession of the Corpus Christi, Sevillefrom the Cruz del Campo, Moorish Tower called theGiralda, Entrance to the Court of the Orange Trees,Entrance to the Hall of the Ambassadors of the Al-czar, the Golden Tower, and the Bull Ring. In 1837another engraving of the interior of the Cathedral waspublished. And in 1837 four lithographs: two of theHigh Altar of the Catedral (pl. II y III), the Giralda (pl.XIV) (fig.9) and a bullfight (pl. XXVI). Roberts paint-ed two oils in Seville, today in Downside Abbey, aninterior of the Catedral and another of the Giralda.There is a water-colour of a bullfight in the Maes-tranza in Windsor37; a view from the Cruz del Campoin Leeds; an interior of the Catedral in Birmingham; agouache of the Torre del Oro (Golden Tower) in NewYork38; and an oil of the same subject in the Prado.In the surroundings of Seville, Roberts made a sketchof the Ruins of Italica, later coverted into an engrav-ing (1836), and a drawing in pencil and a water-colouris preserved of the castle of Alcal de Guadaira39(1833) which would result in a spectacular oil-paint-ing of a sunset, today in the Prado (fig.10). There arealso sketches of Two Spanish Peasants of the sametown (1833), in Denver Art Museum, similar to othersof the Royal Academy of Arts. For Carmona there isan engraving of the entrance to the town (1836), basedon a water-colour with another sunset (1833) whichis today in the Museum of Fine Arts in Oviedo; thereis evidence of a similar oil-painting (1853) alreadymentioned, which Roberts gave to his dentist. Thereare also two lithographs (1838): on the chapel of theVirgin (pl XV) and the marketplace (pl. XVII), a priorsketch of which is preserved in Dudley Museum.There is also water-colour of the Alczar de Carmonain the Victoria and Albert Museum40.While he was in Seville, Roberts had to leave thecountry hastily, as his Record Books indicate: thecholera epidemic was declared in Seville, and thisput an end to all my activities. I intended then to goback and take up again my journey through Spain,but a cordon having been thrown up, I succeededwith great difficulty in obtaining a passage aboard anEnglish ship. After a long and troublesome crossingof five weeks I arrived in Falmouth at the end of Oc-tober of 1833.

4. Views based on views by other artistsGiven that Roberts was unable to complete his jour-ney, for the third and fourth volumes of of The Touristin Spain he decided to use other travellers sketch-es of places he had not visited himself, described inwhat follows. It should be borne in mind that Robertsname appears beneath all the engravings in VolumeIII; but there are serous doubts in three instances:Segovia, El Escorial y Toledo. Although he might have

in a book: whoever is observed sacando planes, tak-ing plans, mapeando el pais, mapping the country, for such are the expressions of the simplest pencilsketch is thought to be an engineer, a spy, and, atall events, to be about no good. The lower classes,like the Orientals, attach a vague mysterious notionto these, to them unintelligible, proceedings; whoev-er is seen at work is immediately reported to the civ-il and military authorities, and, in fact, in out-of-the-way places, whenever an unknown person arrives,from the rarity of the occurrence, he is the observedof all observers.28.Roberts left Granada for Mlaga on 25 or 26 Febru-ary, and must have passed through Loja, since thereis an engraving of the town and its Alcazaba as wellas an original drawing29. Of Mlaga there are twoegravings (1836): a view from the Moorish Fortressof Gibralfaro, and another of the Cathedral; also a lith-ograph (1837) of the fortress and the port (pl. XIX). Healso produced an engraving of the English Cemetery30.On his way to Gibraltar, Roberts passed through Ron-da, whose mountain setting Roberts said was one ofthe most extraordinary in Spain and the world. He de-picted it in two engravings (1835): a general view ofthe walled town and another of its remarkable bridgeover the gorge (Tajo). A watercolour antecedent tothe general view (1834) is preserved in the TateGallery31 and an oil (c. 1835) from a similar perspec-tive in the Huntington Library and Art Gallery, Cali-fornia. Another engraving (1836) includes the townof Gaucn looking towards Gibraltar.Having arrived in Gibraltar on 19 or 20 March, Robertscrossed the Strait to visit Tangiers, Tetun and othercities of North Africa. On his return to Gibraltar, heremarks in another of his letters that the governorinsists on taking charge of my drawings and send-ing them direct to England to avoid the risk of theirpassing through Spain, and this offer I gratefully ac-cepted as my sketches now exceed two hundred32.In Gibraltar he produced a vignette converted into anengraving (1836) and a lithograph (1837) of Gibraltarfrom the Neutral Ground (pl. XXV) (fig.8). There arealso references to an oil-painting of Gibraltar fromRonda (1853)33. Roberts set off in the direction of Cdiz on 23 April;there is an engraving of Tarifa (1836) and another ofCdiz (1836). There are three engravings of Jerez (1836),Xerez, from the Ramparts, the Interior of San Miguel,and Monastery of the Cartuja; also two lithographs(1837), the Interior of San Miguel (pl. IV) and anotherof the church of Santiago (pl. VI). There is a drawingof the entrance to Jerez in Agnew and Sons34; an orig-inal of views of San Miguel in the British Museum35; al-so a drawing of the tower of the Atalaya attached toSan Dionisio in the National Gallery of Scotland.At the beginning of May Roberts arrived in Seville,where he remained five months. There he met young

40 / QUESADA, L.: Pintores espaoles y extranjeros enAndaluca, p. 88, 1996.41 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 109-115,2002.42 / I have made many sketches which are perhaps notup to much, but they are true to reality of the majority ofthe places you have not visited, which you may use to makeyour own drawings. GIMNEZ CRUZ, A.: La Espaapintorescap. 342-343, 2002.43 / ROSCOE, T. (text) / ROBERTS, D.: The tourist in Spain,prefacio, p. V, 1838.44 / SIM, K.: David Roberts R. A., 1796-1864. A Biography, p.64, 1984.

34 / QUESADA, L.: Pintores espaoles y extranjeros enAndaluca, p. 88, 1996. 35 / Reproduced in the central pages of SIM, K.: DavidRoberts R. A., 1796-1864. A Biography, 1984.36 / ARIAS ANGLS, E.: El paisajista romntico JenaroPrez Villaamil, C.S.I.C., Madrid 198637 / CATLOGO: La Sevilla de Richard Ford 1830-1833, p. 64-65, 69, 288 (cat. 123, 125, 126), 2007.38 / VARIOS: Iconografa de Sevilla, 1790-1868, p. 159, 178,188, 189, El Viso 1991.39 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 107(cat. 80), 1986.

45 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 56 y110 (cat. 98, 99), 1986.46 / SIM, K.: David Roberts R. A., 1796-1864. A Biography, p.63 y ss., 1984.47 / GUITERMAN, H. / LLEWELLYN, B.: David Roberts, p. 50 y111 (cat. 106), 1986.48 / Reproduced in CATLOGO: La Sevilla de Richard Ford1830-1833, p.39, 2007.49 / Drawings in the possession of the Ford family.50 / Reproduced from CATLOGO: La Sevilla de RichardFord 1830-1833, p. 40, 2007. Roberts drawing of Santiago deCompostela is held in The Wallace Collection.

206

207

51 / GIMNEZ CRUZ, A.: La Espaa pintorescap. 346-348,2002.52 / Letter of 2 July 1833 from Seville. Quoted in GIMNEZCRUZ, A.: La Espaa pintorescap. 273, 2002.

1 / Beautiful composure of this lower and variedarchitecture, which, between shadows and distance,usurps these heavenly reflections.

it still exists, even if it is a little strange. We refer tostrict, sober architecture, which keeps its spells (andit has many of these among its resources) to itself,in the warmth of its peace-loving interior and in thesecret whisperings of its contented neighbourhoods. This is the architecture that rescues the ground onwhich it leaves its trace and restores it, but not onthe roof, as did the master, but at the appropriateheight, where those who live above meet those wholive below, and to which public spaces are nostrangers. The residents of the upper floors are in themajority (there are more floors above), but those wholive below are no fewer, since the next floor down in-cludes passers-by and neighbours. And that whichis common to them is neither up nor down, at the be-ginning or at the end, but halfway up, at the cross-roads which acts as a neighbourhood landing. Nei-ther too high (because they are not looking to admirea view) nor too low (because neither do they wish tomeet at ground level). This collective third floor, between the second andthe other private ones, stands out, as is only natural,in the geometry of the block. It bites, but not too hard.This is the effect that the common good always hason individuals when they move into the heart of wellchosen company. Other services, the laundry and its surroundings, notshared, but connected, promote contact betweenthose who live on the same floor and invite to play inthe corridors. These are semi-private spaces that en-courage community living in the abstract domainknown to lawyers as horizontal property. With these and other services, a whole communityof people, but not too many, organised with commonsense but also with their own individual sense (bothof which are indispensable) lives, moves and existsin an economic space (in the real sense of being sub-ject to domestic and social norms that rule out bothwant and waste) to which architecture fits, in eachcase, the discreet scale which it knows better thananything else in the world. We are not in either aclosely restricted space or in the unlimited spaceof those who suffer from delusions of grandeur. Wehave, after all, to live within our means. Thus, the block of officially protected flats, abstractin the best sense, that is to say concise, satisfies fromwithin modern discipline the classical Albertian ide-al of the conncinnitas, which Lozano translates ascomposure (Caldern): Let nothing be in excess andnothing lack and everything be in its place. Which,in surroundings with mad figures and dreams of com-mercial fantasy is provocative (like Ulysses withinrange of the Sirens, or the ardent Baptist in the pres-ence of the chaste Salome). The zeal of this quiet anddiscreet architecture, measured and pondered, sim-ply modern, stokes the fires of the post-modern im-ages that dispute its name.

THE OTHERby Joaqun Arnau Amo, Elia Gutirrez Mozo

Hermosa composturaDesta inferior y varia arquitectura Que, entre sombras y lejos, A la celeste usurpas los reflejos.1

Pedro Caldern de la Barca.

The work of Javier Garca Solera is impregnated andsustained with a sort of basso continuo: or beauti-ful composure. This beautiful word, which in Span-ish describes the difficult balance between posture,which is something natural, and composition, whichis inevitably artificial. The word connotes a polite,friendly, serene and graceful manner together witha firm, determined and very personal, highly unusu-al and almost arrogant posture. This beautiful com-posure of architectural technique is founded on theimpeccable execution of this architect (from whomwe have been led to expect nothing less) whose artis expressed in the form of buildings. A combination of elegance and attention to detail isa constant element in the work of this architect,whose career has included many awards, the lat-est being from the 10th Spanish Architecture and Ur-banism Biennial for the building of 40 apartments un-der the Official Protection of the State. In surroundings such as those of Benidorm, a sort ofwarriors garden, a parody of that which coversGauds Pedrera, there suddenly appears a UFO, sup-ported only by clouds and the sun, with a little dis-turbed earth at its base with the occasional bush inthe garden that is soon bound to appear. And he, therara avis, gratefully responds to the elements withnoisy metallic reflections and opaque concrete si-lences, in a perfect replica of Calderons metaphor. But the others, the warriors, assault it. Because theyare warlike and because they are where they are.They are like the forces (in war, as in war, and in themarket as in the market) that impose a (false) peaceby war. They threaten and are ridiculously arrogant.They all have different gestures, but all display thesame aggressiveness that is magnified by their omi-nous shadows. And on viewing this city-fiction, thenewly arrived traveller perhaps is reminded of thatredeeming parallelepiped which, in 2001 SpaceOdyssey, was a symbol of human reason. By the geometry in the sand of the apparently desertedisland, Vitruvius has the shipwrecked Aristippusrecognise that intelligent beings live in that place. Ifthey are geometricians, Aristippus says to himself,the inhabitants must be human (it was Euclideangeometry and of course not fractal geometry). Todaystraveller now remembers that there was once a civil-isation, or perhaps more than one, who knows? Andthere was architecture to sustain it. The proof is that

his own choice of engravings to text by ThomasRoscoe, he received ?420. According to Gimnez Cruz,with the exception of Turner, this would be the high-est price paid in the period for drawings of this type.It was a similar story with the lithographs publishedin 1837, which Roberts supervised and perfected oversix months. After their publication, 200,000 copieswere sold and they were still being printed twentyyears later51. His views were the object of countlessreeditions and imitations. This phenomenon meant that many Spanish citiesemerged as tourist landscapes, drawn and dissem-inated among an ever-expanding public. Thus, beforethe arrival of photography illustrations put on the mapareas of Spain which had until then been remote andexotic. The world became smaller and more acces-sible from the parlours of European middle-classhouseholds. Even today no respectable engravingshop would be without its stock of Roberts lithographsor engravings.In 1838 Roberts embarked on the second great ad-venture of his life, his famous trip to Egypt and theHoly Land, from which he returned in 1839 with 272sketches and drawings, which were to serve him inthe decade to come as the basis for his passion-ate and sophisticated work. Between 1842 and 1849six volumes were published entitled The Holy Land,Syria, Idumea, Arabia, Egypt and Nubia, with a to-tal of 247 lithographs magnificently executed byLouis Haghe. Out of his conviction that the effects of light andshadow [are] all that is great in art52, David Robertswould reach the height of his career in 1841 when hewas elected member of the Royal Academy, the high-est accolade to which a British painter could aspire.Roberts continued to work until his death in 1864.

INDEX OF ILLUSTRATIONS:Fig.1. Portrait of David Roberts (lithograph) by Charles Baug-niet, 1844.Fig.2. Unidentified land (pencil drawing). Private coll. D. LuisLara, Sevilla.Fig.3. The Cathedral of Burgos (engraving). Authors privatecollection.Fig.4. View Toledo (engraving). Authors private collection.

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

YOU OBVIOUSLY BELIEVE IN ARCHITECTURE, SINCE YOU CRE-ATE IT. APART FROM YOURSELF AND A FEW MORE, WHOELSE IS INTERESTED IN WHAT WE PRODUCE?

I think Architecture as an activity that seeks to es-tablish links with real life is of interest to more peoplethan we think. They probably would not say so, butthey are happy with what we do. The problem aris-es when we use architecture to speak about ourselvesand establish a different identity to others. When weartistise it, then we create an abyss between thearchitect and everybody else. It is probably true to saythat what I would call architecture is of no interest ei-ther to those in power or to big business.

IN ANY CASE, TO WHAT DO YOU ATTRIBUTE THE NECESSARYCREDIBILITY OF OUR WORK, OR ON WHAT SHOULD IT BEBASED, IN ORDER TO CARRY ON WITHOUT PROBLEMS?

On thinking that every opportunity that turns up is theone we have and is real; on the unconditional posi-tion of defending the dignity of this work and of every-body involved in it; on the commitment we have (orshould have) to give an authentic social service; onknowing that it is not we who should be talking aboutour architecture and on having the necessarystrength to say no and to carry on without being dis-couraged.

YOU HAVE WRITTEN THAT YOUR INTEREST AS AN ARCHI-TECT HAS CHANGED WITH TIME FROM DRAWING TO BUILD-ING, IMPECCABLE IN YOUR PROJECTS. IN AN IMAGE-BASEDCULTURE LIKE OURS, WHICH IS SUPERFICIAL BY DEFINITION,HOW IS THIS STRATEGY TURNING OUT?

Architecture is nothing without the physical realityof building. Drawing is and always has been very use-ful to us in our continual learning process but meansnothing to all the others who are the final consumersof all architecture. Oza said that the best architectsdrawing is the one given to the worker to help himcarry out a job. I am convinced, and I know what I amtalking about because I have had the experience, thatevery job whispers into the architects ear how itwants to be done. Building can live without archi-tecture but architecture cannot live without building.

HOW DO YOU SEE THE PROFILE OF THOSE WHO ARE THEUSERS IN THIS CASE (PROTECTED APARTMENTS INBENIDORM) OF YOUR ARCHITECTURE? HOW DO YOU PRO-TECT, IN FACT, THE ELDERLY WHO LIVE IN THESE APART-MENTS? CAN ARCHITECTURE CONTRIBUTE TO THE CARE OFOTHERS WITH TOTAL RESPECT FOR THEIR INDEPENDENCEAND WAY OF LIFE? SHOULD IT DO SO?

I have always said at the best thing about private jobsis to build for somebody in particular and that thegreat and beautiful responsibility of public jobs is toimagine that person and work for him. To get closeto the life, see without being seen, get involved andunderstand, try to understandanything that willbring you closer to an answer that contains the many

This architecture is solid but seems light, strong, butseems fragile. It is as firm as an oak but as elegantas a reed. Its severity hides its delicate elements,which expand it. Its reason is the outer coat that pro-tects dialogue, the only activity safe from fanaticaloutbursts. And here they are, one and the other, rea-son and dialogue, constructed and (we could say ifthe word did not sound like the cheap morality of an-other age) edifying. They are not so in this outdatedsense that wishes to see architecture as a school ofgood customs. But they are so inasmuch as they con-sent, and even protect, a certain life in common. And a final note. Protection (tutela) and protec-tor (tutor) are words that now seem confined to thefield of learning and which maybe we should makemore use of. Both are from Latin, in which the matrixof the family is a verb, tutari, in turn derived from an-other, tueor, which means to see in the sense of tocare for, to pay attention to, to be alert to ..

The 40 protected apartments for the elderly in the LaCala district of Benidorm, Alicante, were awardedthe Officially Protected Housing Prize in the 10thSpanish Biennial of Architecture and Urbanism.

Date of Project: September 2005.Work finished : September 2008. Architect: Javier Garca Solera. Collaborators: Lola Prez Pay (architect). Domingo Sepulcre (structure). Quantity surveyors: Marcos Gallud and Javier Mateu. Promoter: Instituto Valenciano de la Vivienda. Construction Company: Sedesa Obras y Servicios S. A.Total Budget: 3,085,892 euros. Constructed area: Buildings: 5,240 m2. Urbanisation: 1588 m2. Cost per sq. metre: 589 euros/m2.

INTERVIEW WITH JAVIER GARCA SOLERA

WITHOUT GOING INTO THE USUAL CLICHS (CRISIS, SUS-TAINABILITY, ETC) BUT WITHOUT IGNORING THE BURNINGQUESTIONS, IN YOUR OPINION, WHERE IS ARCHITECTUREHEADED IN THESE TIMES OF ECONOMIC RECESSION?

It is difficult to say what future projects will be like,those which perhaps will be derived from a processof reflection due to this general crisis. To be honest,I think that the view that will be heard most will againbe the most superficial or falsely committed, whichwill know how to handle the situation in order to oncemore generate surprising architecture easy to sell. Ithink that the sheer manipulation that is constantlymade of terms such as sustainability by architectsthat seek refuge under its protection provides suffi-cient grounds for my doubts.

Because the warriors are there and do not seem dis-posed to leave the field, in their frenzied and grotesquetribal dance. Only that, as in Amenbars films, thosethat seem to be always there, installed in perpetuity,are The Others. They, and not the one supposed to beout of the ordinary, are the UFOs. They are the new-comers from another planet (perhaps uninhabitable).They dazzle our eyes and shout at us. And meanwhileamong them there is real architecture that is almostinvisible (which is what we are at present dealing with),which takes up its ration of space, no more than itneeds, and manages it for the benefit of its protectedbuildings. Because it itself is the protector. Its arrangements serve but are not servile (thecaprice that perverts the right use of liberty have beenbanished from it: everything in this building is coher-ent and thought out). In their own way they providethe best and the most silent protection. Careful ar-chitecture that cares: refined and behaves with re-finement. And its protection proposes peace as a premise: apeace that stands out by not standing out, that at-tracts attention by not attracting attention, in the jun-gle, not the garden, of the warriors. It is earthly, in aplace that puts on interplanetary airs, real among theshadows, active and alive among richly decoratedmausoleums. However, after a long quiet talk in themidday sun, the landscape we see is transfigured.The urban scene (can we really call it that? Or is itco(s)mic?) moves on (as in the old time theatre). Thewarriors disappear into the black holes, taken awayin the mother ship. We are at home and we are not alone. There are neighbours that protect themselves fromthe sunlight, poke their heads out, go in and out, wan-der around. They think, perhaps, that the others watchthem, because they do not know that although theyseem to be there they are not really there. Neither dothey know, who knows? that the sockets of the oth-er eyes are empty. They did not come to see, onlyto be seen. And they are like smoke, that can be dis-pelled by the slightest sea breeze. As Loos would say,this, and not the other, is architecture. But he saysit from a simple tomb, which this is not, although itappears to be simple. This is the house, common andindividual, turned into the protector of a common life. A certain tradition has it that architecture descendsfrom the sedentary crop-planter Cain, fratricidal broth-er of the shepherd and nomad, Abel. The architect,who builds to be permanent, belongs to the race ofCain. But the architect Garca Solera turns aroundthis sinister tradition, which is not very favourable tothe profession. To the question, Am I my brotherskeeper? he firmly answers Yes. We all look aftereach other in an architecture, which, for a start, looksafter us. And when one is older, this is not only de-sirable but essential to a fuller life.

208

209RESTORING, REDRAWING, VENTURING.STRATEGIES FOR GRAPHICALLY DOCU-MENTING THREE MONUMENTAL GATE-WAYS OF MADRID

by Aitor Goitia Cruz

The venerable exercise of surveying presents thedouble fascination of being imperishable as well asin constant renovation. As ancient as the need forlearning and as personal as the interests or abili-ties of those who practice this art, its utility and beau-ty have played a leading role in essential moments inthe history of architectural drawing. Andrea Palla-dio, Paul Letarouilly, or Luigi Canina to mention on-ly some of the most famous masters of documentaldrawing have allowed us to know not only the formsof older architectures, but to also understand theircomposite and tectonic keys. All of this is possible thanks to the mediation of draw-ing as an indispensable instrument in attaining andtransmitting knowledge. This perceptive aspect ofgraphic expression is found present in other appli-cations of architectural drawing, but, without doubt,is exercised in each phase of elaboration of an ar-chitectural survey: from the first sketches to the fi-nal drawings, through gathering data, general con-structions, analyses, comparisons or fine adjustmentsof forms and dimensions. In recent times, certain sci-entific and technical advances have completely trans-formed the manner of obtaining and processing da-ta extracted from constructed works. The traditionalrules, tape measures, and plum lines, are now com-plemented- not substituted- by analogical and digi-tal land photogrammetry, total stations, or by 3D laserscanners. However, independent of the availabilityof these means, the principal task of the draughts-man resides, as long ago, in the adequate interpre-tation of the architectural work chosen and the cor-rect manipulation of the data obtained, in order tofinally determine a graphic model analogous to thetectonic one. For this reason, whichever the objec-tives, the models chosen, the techniques of inspec-tion and measurement, or the graphic solutions adopt-ed, the interpretative quality of architectonic drawingrequires a conscious investigative attitude in the ex-ploration and determination of the elements com-posing the object of the study. As an example of practical applications of these tech-niques and attitudes, I present three surveys1 of threemonumental gates in Madrid, which are also the ob-ject of other deeper and longer studies than those in-cluded here. These are the three gateways from Tole-do, Alcal and Atocha, which correspond to situationsdifferent from the initial data. I also discuss the meanswe disposed of to approach each of these graphicdeterminations. These gateways were chosen as the

tional performance. In Benidorm, all we did was toexplore the vertical repetition of the many variationsof this arrangement.

BESIDES THE GROUND AND FIRST FLOORS, YOU ALSO IN-CLUDED THE THIRD FLOOR FOR COMMON USE: THE SECONDAND THE REST OF THE FLOORS CONTAIN THE 40 APART-MENTS. WAS THERE A PARTICULAR REASON FOR PRECISE-LY THIS FLOOR, AND NOT ANOTHER, TO INTERRUPT THE VER-TICAL CONTINUITY OF THE BLOCK?

This floor maintains a very intense relationship withthe space of the street, which reinforces its publiccharacter. The fact of being between floors of apart-ments gives the feeling of being inside a private areaand so helps to welcome those who enter in com-pany with those who live there.

ORTHOGONAL GEOMETRY GOVERNS ALL PROJECTS WITHABSOLUTE PRECISION, HOWEVER, THERE IS SOMETHING INTHE BUILDING THAT STRIKES US AS HIGHLY SIGNIFICANT.WHY ARE THE HANDRAILS INCLINED AT AN ANGLE?

A close look at the section will explain how thesehandrails try to overcome the condition of forminga limit, which they normally do. It is an invitation tosit down facing inwards using the handrail as a sup-port with part of the body on the outside. Also, thisconfiguration will help people when tending theirplants which we expect to see on the metallic traysthat act as sunshades over the windows on the front.

THE LAUNDRY SPACE RECEIVED SPECIAL TREATMENT.COULD YOU TELL US HOW IT IS ARRANGED?

The laundry is a communal service, so what betterthan to put it on this floor where people can meet, sitout in the fresh air and talk while they are waiting forthe washing cycle to finish.

AMONG YOUR PROFESSIONAL COLLEAGUES WHO ARE STILLACTIVE, FAMOUS OR NOT, IS THERE ANYONE YOU PARTIC-ULARLY ADMIRE OR WHOSE CONVICTIONS YOU SHARE?

I am always more interested in the work than the ar-chitect. Some almost always have something to teachme: the simplicity of Kazuyo Sejima, the economy andtechnical precision of Lacaton-Vassal, the identifi-cation of architecture, construction and structure ofPaulo Mendes da Rocha..

WHAT WOULD YOU SAY TO THE YOUNG, OR NOT SO YOUNG,STUDENT WHO HAS JUST RECEIVED A GOOD GRADE FORHIS FINAL PROJECT?

You are an architect, nothing can be indifferent toyou. And do not be in any hurry.

La traduccin de este trabajo ha sido financiada por la Uni-versidad Politcnica de Valencia.

The translation of this paper was funded by the Universi-dad Politcnica de Valencia, Spain.

ways of occupying, of living in that space. All I do istry to offer possibilities in such a way that architec-ture is present but is never the protagonist.

YOU HAVE SAID THAT YOU DO NOT WANT TO ACCEPT JOBSOUTSIDE YOUR RADIUS OF ACTION SO THAT YOU CAN PAYFREQUENT VISITS TO THE WORK SITE. BENIDORM OBVI-OUSLY COMPLIES WITH THIS REQUISITE, BUT COULD YOUTELL US IF THIS DECISION HAS LED TO YOU HAVING TO RE-FUSE IMPORTANT CONTRACTS? IN THE GLOBAL VILLAGE,DO YOU NOT FEEL TOO CONFINED TO ONE LOCALITY? ANDIF YOU CAN TRAVEL TO GIVE LECTURES, WHY NOT FOR REA-SONS OF WORK?

It happens that I need to be pretty close to the build-ing work in order to hear the latest whispers from thejob telling me how it wants to be done. This means Itry to look for work within a radius of action that al-lows me to go there frequently. I do not actually lookfor work outside this distance, but this does not meanthat I would refuse a contract a long way away. If thecase arose, I would make sure that the distance wasno obstacle to continuous visits.

WE HAVE ALSO HEARD YOU SAY THAT YOU CARRY OUTPROJECTS AGAINST CIRCUMSTANCES. IN THE CASE OFBENIDORM, WHAT CIRCUMSTANCES DID YOU HAVE TOFIGHT AGAINST?

In Benidorm we had to work against the conventionalidea that public and private jobs are different, thatthe city is only drawn on the ground, that limits haveto be physical and rigid , that the common practiceis not possible, with the old with the old and the youngamong them, that the street is a frontier, that con-crete is a hard, cold material, that metal is hard andcold, that a building has a back and a front, etc.

YOUR PROPOSAL FOR THE PROJECT INCLUDED A TOWER (FORTHE YOUNG) AND A BLOCK (FOR THE OLD PEOPLE). DO YOUTHINK THE DEFINITIVE SOLUTION OF THE BLOCK WITHOUTTHE TOWER HAS BEEN DETRIMENTAL TO THE PROJECT?

Benidorm has become quite famous but for me it willalways be a frustration. The wide recognition that ithas received reassures me that it has been a suc-cess, but I think that the complete project had somuch in social and urban terms, technical possibili-ties, I mean in architectural terms it could havebeen much wider. It pains me to think we were onlyone step away2.

YOU USUALLY WORK WITH VOLUMES NO GREATER THANTHREE STOREYS. IT COULD EVEN BE SAID THAT A MIESIANSPACE OF A SINGLE FLOOR IS YOUR NATURAL MEDIUM INWHICH YOU SEEM TO FEEL AT HOME. HOW DID YOU TACK-LE THE 11-STOREY BLOCK IN BENIDORM?

It is true that I feel at home with fluid, horizontal spaceand also with the efficacy of two planes (floor andceiling) to create valid interior-exterior and natural-artificial relationships, apart from their excellent func-

ex

pre

si

n g

rfic

a a

rqu

itec

tn

ica

2 / The Benidorm project was based on two competitionsfor two sites on different sides of the same street, one ahome for the elderly and the other for young people. Weproposed a joint project for both sites that even had abridge to join them. We obtained the first place for the oldpeoples residence and the second for young people.

1 / In a strict sense, we should not speak of survey in all ofthe cases, as one of the edifications studied no longer exists.

ments using a measuring tape, which, given the sizeof the work, hardly led to the true knowledge of thelower base and the immediate area of implantationin the roundabout. The lack of scaffolding or auxil-iary structures which would allow continuing withdirect measurements of the construction, obligedus to use a photogrammetric restorer10. The nor-mal method in the use of this instrument, - pairs ofphotographs taken in situ under the necessaryconditions of distance and convergence, adaptedto the difficult conditions of surrounding traffic, wasemployed. It is necessary to specify, however, thatthe photographs used for the three-dimensionalrestitution of the model were obtained with a con-ventional a 35 mm-negative reflex camera. Althoughit is not the most precise method that this techniquemight facilitate, we must point out that it was the on-ly one possible at the moment of the execution ofthe survey. Most of the dimensions and the principal distanceswere obtained with the restorer. The proportions ofthe minor elements of the cornices, bases, capitals,etc., were established by the larger dimensions thatfixed them, the photographic information, the cor-relation between metric systems, and the compari-son with Vignolas rules of order of Architecture, withwhich the work of Lopez Aguado systematically co-incides, except for some ornamental details. The drawings elaborated for the definition of the Puer-ta de Toledo were obtained by computer, using Au-toCAD. The reference scale for the paper edition wasestimated at 1:50, so that the formal details of the dif-ferent elements of smaller dimensions could be ap-preciated. The drawings which accompany this textare adjusted reproductions, and therefore, the rela-tionship between the size of the drawings and theirdefinition are not appropriate. Grouped in orderedsets, the following documents were produced:

Elevation facing the exterior of the city. Ground planThe elevation towards the exterior of the city is moreornamental, bearing in mind the apparent austerityof the gate. Its graphic representation has also beenconsidered as the most emblematic image amongthose obtained in this survey, therefore deserving acomplete treatment, more elaborate than the rest.We included all of the lines of projection that restorethe forms of the unit; the arrangement of the mason-ry work; and the most impeccable image of the sculp-tural works of Barba and Salvatierra, as this is thefaade in which they offer all their symbolic impact.The ground plan accompanying the main elevationclearly transmits the different composition of thefaades towards the interior and exterior of the cityby means of a group of columns and pilasters.

Elevation facing the city. Ground planThe secondary elevation, frankly, is something more

lowing year. The prolonged history of the construc-tion of the gateway is replete with economic diffi-culties and political upheavals - the return of Fer-nando VII, Riegos pronouncement, a liberalthree-year period, an ominous decade - which thegate is witness to, and also object of continuous spec-ulations and changes in the documental databasesand dedications. Once completed, the constructionwould continue to undergo new manipulations andproposals, such as its demolition, solicited by vari-ous councilmen who maintained this initiative7 be-tween 1879 and 1881.

Existing graphic documentation. Strategy adoptedIn spite of the importance of this work, its enclaveand historical significance, the search for graphicdocumentation relative to the Puerta de Toledo sit-uates us before a most singular case. The disap-pearance of Aguados drawings, contained in therecords8 of his proposal, and the total absence ofreferences to their possible location, suppose an ir-reparable loss. This misfortune is repeated with theplan that the architect elaborated in 1824 for the ex-ecution of the registers and modification of align-ments9 of the ensemble. However, drawings and en-gravings in which the monument appear areabundant, although lacking adequate fidelity to thereality of the gateway, as well as to its definition asto ground plans and elevations. On the other hand,plans carried out for other entrances executed inthe city have been conserved: along with the afore-mentioned royal gateways of Alcal, Segovia, andFuencarral, ornamental gateways such as Recole-tos and San Vicente or the gates of Gilimn, Emba-jadores, or Valencia count on original or redrawnplans. It is only the absence of similar documenta-tion which equates the case of Atocha to that of thePuerta de Toledo. As a consequence of such a lack of documentation,approaching the survey of the Puerta de Toledo sup-posed, in a certain way, a compromise with history.With the aid of only the monument itself, the strat-egy and the procedure adopted have been directedtowards the graphic determination of the originalstate of the monument, dispensing with the registryof associated deformations or pathologies, moreproper to studies leading to architectural interven-tion. Therefore, certain defects in form or fracturesthat some parts present have not been borne inmind, as we have opted for the regularization of theelements that intervene in the monument, lookingfor a certain dimensional and formal idealizationwhich reflects the supposed state projected by An-tonio Lpez Aguado.

Gathering data and graphic constructionThe work involved in the gathering of data began witha direct measurement of all of the accessible ele-

object of study because of their singular monu-mentality and due to a lack of trustworthy graphicdocumentation, which would permit both under-standing them individually, as well as their com-parative analysis. These three gateways were located along the wallthat withheld the growth of the city for more than 200years2 and were situated in strategic points of thecity, as they led to three of the five so-called royalroads. The other two roads were presided over bythe gateways from Segovia, to the west, and Fuen-carral, to the north of the city. Of the latter, which dis-appeared many years ago, the original drawingsmade by their authors, Teodoro Ardemans3 and JuanGmez de Mora4 are conserved today. To the con-trary, neither the drawings that gave birth to the Tole-do Gate (Puerta de Toledo), nor those that correspondto that of Alcal seem to adjust to the monument thatwas constructed. In the case of the Atocha Gate(Puerta de Atocha), both the absence of drawingsand edification aggravate the lack of documentation,which requires redress. In light of these antecedents, the basic objective ofthe surveys undertaken has been focused on ren-dering new graphic documents which might mitigate,at least in part, the architectonic definition containedin the drawings of the lost projects. Although the in-tention and essence were alike -obtaining data andgraphic elaboration- the peculiarity of each case hasinfluenced the work strategies and the graphic solu-tions, which are set out below.

GATE OF TOLEDO (PUERTA DE TOLEDO)

Some factsThe monument, concluded by Antonio Lpez Agua-do in 1827, would be the last gateway constructed inMadrid. As its surroundings have been totally altered,today the Puerta de Toledo appears isolated and in-comprehensible because it must be viewed from soclose up. This incomprehension extends back to thevery same history of the monument, having been tak-en as a certain false interpretation that historians andchroniclers repeated, attributing the initiative of itsconstruction to Jose I or Fernando VIIwhom it wasfinally dedicated to.The first documentation of the work of Aguado, whichshould have included his design plans5, is a documentdirected to the Ayuntamiento Constitucional on No-vember 13, 1813, by the governor Santiago Gutirrez.He proposes the erection of the new Puerta de Tole-do to welcome the Soberano Congreso Nacional,which would renew in Madrid the sessions whichwere celebrated in Cdiz since its beginning in Sep-tember 1810. The foundation stone was placed on De-cember 22, 1813, months after the definitive flight ofthe French king6, and before the return of FernandoVII, which was not effective until the spring of the fol-

10 / The restorer used was Adam MPS2, along with anAutoCAD 12 application, from the Departamento deIdeacin Grfica de la Escuela Tcnica Superior deArquitectura de la Universidad Politcnica de Madrid..

2 / The construction of the enclosure was initiated in 1625by order of Felipe IV. With diverse modifications in the wallsand gateways, it remained standing until the demolitionsthat began after the approval of the widening projected in1860 by Carlos M de Castro.3 / Archive of Villa de Madrid ASA 1-201-28: Puerta deSegovia. 1, External face: Ground plan and Elevation. 2,Internal face towards Madrid: Elevation, 1703.4 / Archive of Villa de Madrid ASA 1-204-9: Puerta deFuencarral. Ground plan and Elevation, 1642.5 / Archive of Villa de Madrid ASA 1-201-6. Dossier with thedocument of Gutirrez to the Town Hall, which accompaniedthe design of Antonio Lpez Aguado, since lost.

6 / Jos I abandons Madrid on March 17, 1813, enteringFrance on June 27 (after the Battle of Victoria), immediatelyrenouncing the Spanish crown.7 / Archive of Villa de Madrid ASA 6-384-6.8 / Archive of Villa de Madrid ASA 1-201-6. Register cited.9 / Archive of Villa de Madrid ASA 1-201-11.

210

211

11 / The gateway conceived by Sabatini put an end to thesuccession of names, localities and factories thatpopulated the main entry to the Villa along the road fromAlcal: Arco de Santa Mara (Arab wall), Puerta deGuadalajara (Christian wall), Puerta del Sol (wallsurrounding the poorer quarters), Puerta del Sol-Alcal(Wall of Felipe II, 1566), Puerta de Alcal (1580, 1599 -Patricio Cajs-, 1636, 1691 -Teodoro Ardemans-, 1702).

12 / National Archives of Pars, series NN 23. This albumcontains, among others, four sheets relative to the Puertade Alcal, with Ground plan and Elevation on the sidefacing the city, Ground plan and Elevation on the sidefacing the exterior, Ground plan, Elevation and Transversalsection, and a view of the perspective.13 / The works were projected and directed by Po GarcaEscudero. Thanks to the intervention of Enrique Echevarra,I was given a few reduced copies of the survey, which hadbeen drawn by the draughtsman Francisco Sebastin.

vergence exists opens the doors to new investiga-tions of the records of its construction14 or to more orless risky conjectures. Regarding this, I dare to sug-gest that, given that the drawings in Paris were sent-and most probably re-drawn- on dates very close tothe conclusion of the gate, perhaps Sabatini himselfwished to dote this work with the magnitude that it islacking and the slenderness that many engravers fal-sified, to enlarge the relative modesty of its presencein the most important entrance to the capital of Spain.

GATE OF ATOCHA (PUERTA DE ATOCHA)

Some factsAlthough it is one of the main entrances to the city,adjoining such important routes as the Saln del Pra-do and the Tridente de las Delicias, the Puerta deAtocha was recurrently depreciated. Its predeces-sors were only discrete brick gates next to the Castel-lana watercourse -and the sewer- until Pedro de Rib-era succeeded in obtaining (in 1736) the appearancecreated by Joli on one of his well-know visits to thecity.15 Because of its peculiar style, or of the difficultconditions of construction and maintenance, it un-derwent ephemeral interventions, projects not exe-cuted, and a long list of attentions, with the object ofimproving its aspect. Fernando VII, on his successiveentrances to the city, would be the main promoter ofthese acts, and Antonio Lpez Aguado, the architectthat would give form to royal desires.16

Other triumphal entries through this gateway woulddemand new reformations, such as those carried outin 1828 and 1829 by Francisco Javier de Maritegui,who felt that the gateway was in the most beautifuland visible location of the Court. Later projects of Cus-todio Moreno (1830) or Snchez Pescador (1844)could not be converted into reality. In June 1850, dem-olition was begun, as its construction was to be sub-stituted by a barrier under the line of the railway dock.This plan was never carried out, either.

Existing graphic documentation. Strategy adoptedHaving disappeared one hundred and sixty years ago,the possibility of recreating the image of the Puertade Atocha now solely depended on the existinggraphic documentation, although the low esteem forthis gate seemed to affect artists of that time, whoignored it, even when they depicted its surroundings.This Puerta is, with considerable difference, the roy-al gateway to the Villa with the fewest illustrationsreferring to its construction. Only two exceptions al-lowed us to imagine its aspect with certain assured-ness: a lithograph drawn and engraved by Camarn17

in 1829, which reflected the immediate state of theinterventions of Maritegui, and another, corre-sponding to the collection of Madrid Artstico, datingfrom approximately 1845, suggesting the permanenceof singular ornamental details.

drawings of Sabatini, whose definitive design wasreflected in the so-called lbum de Pars12, dated 30March 1777, and dedicated to Prince Kaunitz Riet-berg, to show him his works of embellishment com-pleted or begun in the last few years.At first sight, the general proportions of the finishedwork seem different from those in the drawing, whichseem more slender. To evaluate this probable differ-ence, the survey of the Puerta de Alcal offered theadded possibility of establishing a precise compari-son between the documents obtained and the draw-ings of the Italian architect. A third element of com-parison, the survey undertaken by the EmpresaMunicipal de la Vivienda13 for the restoration workscarried out in 1993-94, could arbitrate the occasion-al divergences sought. Contrary to the case of Tole-do, the fundamental objective of this survey was notto fill any documental void, but rather to evaluate thefidelity of the constructed work with respect to theoriginal plans of the architect.

Gathering data and graphic constructionOnce again, the inalienable physical contact with themonument and its direct measurement needed to becomplemented with advanced methods which themeasuring tape could not undertake. On this occa-sion, with the inestimable help of my workmatesMiguel Alonso y Ana Lpez, readings of key points ofthe gateway with the total station of the Departa-mento de Ideacin Grfica of the ESTAM were tak-en. I must also mention the collaboration of Elena Al-colea in the posterior graphic reconstruction.With a practically inconsiderable error of measure-ment, we have constructed the ground plan and mainelevation of the Puerta, with the caution proper tometric relations, proportional between sustainedgathering of data in t